^.';T;^>';i^' ; ^^^ c ^ LAVAS AND SOILS OF THE 'HAWAIIAN ISLANDS. Investigations of the Hawaiian Experiment Station and Laboratories. WALTER MAXWELL Director and Chief Chemist. HONOLULU : HAWAIIAN GAZETTE COMPANY. 1808. Consulate Geneol of the United States, Honolu-lu., H. I. APR 27 1898 LAVAS AND SOILS OF THE HAWAIIAN ISLANDS. INVESTIGATIONS • OF THE HAWAIIAN EXPERIMENT STATION AND LABORATORIES BY WALTER MAXWELL, - - - Director and Chief Chemist Assisted by J. T. CRAWLEY First Assistant Chemist, C. F. ECKART Second Assistant Chemist And E. G. CLARKE, Field Assistant. 1, Origin and Nature of Hawaiian Soils. 2. Availability and Loss of the Elements of Plant Food in Hawaiian Soils. Published bv Order of the Hawaiian Sugar Planters' Association, \* i8q8. 6346 Trustees- and Members of the Haicaiiau Sugar Planters^ As- sociation. Gentlemen:—! hereby submit a report, setting forth the results of investigations of Hawaiian soils. The scope of the investigations, involving a careful study of the lavas from which the soils have been formed, may appear, at first sight, to extend beyond the require- ments of the subject: An examination of the report in detail will, I believe, make it very clear that it would not have been possible to arrive at an understanding of the great differences in the nature of the soils, and in their economic values, without a preceding study of the lavaft such as was undertaken. The discussion in detail of the chemical processes, and of the laboratory methods, is required in order that other scientific men may follow the mode of the investi- gations, and judge of the value of the results. In laying the plan of the investigations, and in the study and adoption of methods for its execution, I have endeavored, before all else, to observe and be guided by Nature. For this reason, repeated visits have been made to each plantation and district on the four islands: soils have been examined in place, with the lavas from which they were derived, and in careful connection with the local climatic conditions and environment. In the efforts to establish a reliable mode of estimat- ing the state of availability of the essential element* of plant food in the soils, I have tried to find out the processes, and the results of the processes, that operate in the field; and then to bring the methods and procedure of the laboratory into harmony with these. The result* will be found to amply justify aud reward the course that has been followed. In carryiii^- out such a plan of investigation as I have described, assistance in each part of the work was a necessity. Therefore I wish — First, to acknowledge the aid furnished by the gentlemen on plantations in obtain- ing soils, and in recording the climatic and other local conditions. Further, without the co-operation of the gentlemen in the laboratory the execution of the work had not been possible. To First Assistant Crawley I have been greatly indebted, and not only for the many extremely delicate analytical results that he has furnished, but also foi^ valuable critical observations in the adjustment of the methods of the laboratory, in order to compare with the processes in the field, upon which the laboratory proced- ure was based. Also, in the analysis of lavas and soils. Second Assistant Eckart has rendered indispensable as- sistance. In the outdoor part of the experiments, which Avere conducted in the experiment station field, I have been ably and faithfully assisted, in the carrying out in detail of tests and observations by Field Assistant Clarke. In fact, it is not possible to say too much of the assistance received from these several sources. With these and continued investigations of the soils as a basis, the experiment station is now proceeding with n broad plan, embracing the RcJntioiis; of »s'o//.s' to Crops. Walter Maxwell, Director and Chief Chemist. Honolulu, H. L, 1898. DEFINITIONS. Symbol. Name. Si O, Silica. Ti O2 Titanic Acid. P, O5 Phosphoric Acid. SO3 Sulphuric Acid. CO2 C^arbonic Acid. H CI Hydrochloric Acid. CI Chlorine. Fe O Ferrous Oxide. Fe2 O3 Ferric Oxide. AI2 O3 Alumina. Fe2 AI2 O,; Iron and Alumina. CaO Lime. Mg O Maiiuesia. Mn3 O4 ^Manganese Oxide. K2 O Potash. Nao O Soda. N Nitroceu. ORIGIN AND NATURE OF HAWAIIAN SOILS. The mineral constituents of all soils are furnished by rocks or lavas, as a result of disintegration. This truth, in the examination of particular soils, causes us, in the first place, to give precise attention to the rocks from which these soils have been derived. HAWAIIAN LAVAS AND ROCKS. The Islands of Hawaii are of volcanic origin, therefore the rock materials composing the mountains and eleva- tions above the sea are igneous in character. Professop Dana says, "The Hawaiian Island group is an example of a line of great volcanic mountains. Fifteen volcanoes of the first class have existed, and have been in brilliant action along the line." Petrographically, the lavas and rocks composing the great mass of the structure of these Islands are hasaltio lavas. Eecurring to the definitions of Dana, "these basaltic lavas belong to the same class, although they vary widely: — in color, from dark to light gray; in struc- ture, from compact to highly vesicular, and from those of uniform grain to those which are prominently por- phyritie, with chrysolite or feldspar.'' In his niiueralo- gical examinations, whioli are confirmed by chemical analyses, Dana speaks of only one "remarkable felds- pathic andesyte of a totally different rock from any other obtained from the Islands," which did not "con^ form to that of normal basalt." In r()acli('d tlie location of the active cratei" of 11 Kilauea, a new order of suggestions occurred to him, which, upon a closer survey of the conditions and phenomena presented by the actual crater and its sur- roundings, developed into definite thoughts, and into a series of analytical examinations, which were to bear upon the attempt to explain the causes of difference in Hawaiian soils already spoken of. The crater of Kilauea, according to data furnished to tis by Professor Alexander, Surveyor-General, is 3972 feet above the sea. Within the walls of the crater is enclosed an area of 4.14 sq. miles, or 2650 acres. The present area, or lake of burning liquid activity, may not cover more than fifty acres, and forms a crater within the crater. The total area of the vast crater, however, is covered by activity of volcanic phenomena. Viewed in the morn- ing before sunrise, when the air temperature is low, a large part of the crater floor is more or less concealed by condensing steam, which is issuing from fissures and fractures in the lava, that were produced on partial cool- ing. Some preliminary observations showed us that at some places of escape the steam gave no reaction with litmus paper; in other places a mildly acid reaction was found; whilst over other well-defined areas the issuing steam was intensely acid and hot It was further noticed that, over the steaming areas, deposits were laid of dis- solved materials brought up through fissures; and over areas of intense action of acid steam the general surface of the lavas was undergoing acute change. During the first visit to the crater, typical samples of sound lava, of decomposing lava, and of several decom position products, were taken, and brought to the labor- atory for examination. We made a second visit a few months later, being prepared to make more precise ob- 12 servations ii])(>ii the nature of the active ajjients in lava decomposition actually operating at this time, and to locate and identify some of the products of disintegration. The temperature of the steam leaving the fissures shoAved every degree from the lowest up to the last point of condensation. There are also fissures fi'oni which, doubtless, moisture is escaping, but the temperature is too high for condensation at the lava surface. The writer found no i> commercial i)ni'iH)ses. Over the whole area, sometimes upon the top, in other places a feAV inches beneath the surfaces, and again upon the sides of the fumaroles, masses of pure sulphur are crystallized out. Then, there are alum deposits; heaps of extracted lime, lyinii in the form of almost pure gypsum, and beds of so-called red and yellow ochres, wlierc^in we have to search for the iron and silica. At tliis time, we have before us the most extraor- dinarily illustrative specimens, showing the course of disintegration by Avhicli the solid lava is resolved into tlie products nauKMl. These were obtained by working away several feet of the outer edge of the decomposing lava and securing hloels in place, which had never been expos- ed to the air, and ui)on which the sulphurous steam waft constantly acting. At the tiuje of taking the specimens the lava blocks were so hot that the hand could not touch them, and the steani, on i>assing into the air, was 90°C (194° Fahrenheit), but down in the interior of the lava> in tissures, no steam was visible, showing that tli(^ heat was above 100°C. The water collected by cond(Mising this steam at a point of exit contained 5% of sulphuric acid, but no chlorine. Two of these larger specimen blocks of lava, in states of decomposition, are partially studded with mature crystals of sul])hur. Where the acid steani has eaten out cavities, these are partly tilled with clustered crystals of gypsum, and some alums, each of these being vei-y (U^tinitely sei)arat(Ml. Over the sur- face, and within the blocks is seen the irou gathering to- gether in pockets of red ochre, and the silica separating in distinct masses of a so-called yidlow ochre, and in places the silica is dejxtsited almost pure. With most 17 of the numerous specimens collected by us the separation and crystallization have gone on and matured since they were brought to the laboratory, and we have before us, not only exact, but very beautiful examples of the disin- tegration of the lava, and the formation of the product^^ that result from their decomposition. The action of steam in the disintegration of igneous rocks has been remarked upon by several observers. Darwin, in speaking of steam action upon tracliytic rocks at Terceira, in the Azores, says, ''the steam is emitted from several fissures: it is scentless, soon blackens iron, and is of too high a temperature for the hand to bear." The steam spoken of by Darwin, which he says was "scentless," differs from the acid steam that we have described. That the steam "blackens iron" is suggestive that some acid was present; and as Darwin did not make any analytical examinations, either of the steam or the decomi)osition products, we are inclined to think that he was dealing with phenomena resembling these that we are describing. Moreover, Darwin says "the manner in which the solid trachyte is changed on the borders of these orifices is curious: first, the base becomes earthy, with red freckles, evidently due to the oxydation of particles of iron; then it becomes soft. After the mass is converted into clay, the oxide of iron seems to be removed from some parts, which are left perfectly white, whilst in other neighboring parts, which are of the brightest red color, it seems to be deposited in greater quantity. The inhabitants use these substances for white washing." We are, from these products that Dar- win describes, quite impressed that he was dealing with a corresponding condition of things to what we have ob- served at the Kilauea crater; yet we cannot say more, 18 since the great scientist did not note the presence of definite decomposition i^roducts, and does not fnrnish any analytical data bearing upon phenomena, which he says "are still obscure." Darwin remarks that these things "have been observed in other places, in the Italian volcanic islands, and by Spallanzani, Dolomien, and others," which statement accentuates our persuasion that in speaking of phenomena found at Kilauea we are dealing with matters that have had a more or lesa universal concern in the history of rock disintegration and soil formation on our globe. We shall now give a small series of analyses of decom- position products of lava collected by us at Kilauea. To economize space these are given in a table. A— is a sam]^le of almost pure gypsum which is found distributed in enormous quantities. B — was taken from the "alum deposit," and is a mix- ture of sulphates of the alkalies, iron and alumina, with the excess of sulphur and sulphuric acid. r — is a portion of the so-called "red ochre" which is found in largo masses and layers, and more or less defi- nitely separated from the other products. D — is (•(»m])osed mainly of silica which has been releas- ed by th(^ action of the sulphuric acid in the steam from the lava bases. It is yellow on drying, and becomes pink on glowing. E is a material still showing th(^ lava form, but exhibits complicated modes of disintegration, one result of wln when speaking more definitely of soils. Ill coiiuoctiou with those examinations of the lava, lava decomposition products, and of the agents by which the disintegration is being caused, an experiment was made in our laboratory showing the action of acid steam on lava. Lava, of a known chemical composition, was broken into pieces of the size of a large bean, and put in^ to a glass tube. This tube was connected with an Erlenmeyer flask containing a five per cent, solution of sulphuric acid, which is the acid strength of the con- densed steam operating at the crater Avhere our sample* of the products of decomposition were taken. The other end of the tube was connected with a condenser, by which means the acid solution rose as steam through the lava in the tube, and returned to the flask. Exactly 52.9 grams of lava were put into the tube, and the acid steam acted upon it for 120 days. After this period of action, 1.221 grams of solid matter was found in the solution, the composition of which was as follows, after deducting the amounts of soda and silica dissolved out of the glass of the Erlenmeyer flask. Si O 2 =16.00 per cent. Feo 0.3= 1.70 per cent. AL, 0.5= 4.56 i^er cent. Ca"o= 6.53 per cent. K.> 0= 5.58 iier cent. Nag 0= 5.00 per cent. SO^,, etc.=60.73 per cent. 21 These data are highly instructive in indicating the mode in which the disintegration may be proceeding in nature. Tliey show the amount of silica that is released by the action of the sulphuric acid on the bases in the lava. Also it is seen how the "alum deposits" are formed by the separation of the alumina and alkalies, as suljihates, from the lava. The removal of the lime ac^ counts just as simply for the deposits of gypsum, whilst the iron is less affected, which suggests that those decom- position products of the lava that are extremely rich in iron are residual, rather than separation products, show- ing what is left of the original lava after the soluble silica, and the elements which form the alums and gyp- sum have been removed. There are other modes of disin- tegration that are not yet as well understood, and which result in the evident removal of the iron. The time that has been given to the study of phenomena that are actually visible at the present time in the pro cesses of disintegration operating at the Kilauea crater is for the purpose of determining, if possible, a connec- tion between what is now going on at the volcano, and what may have taken place in other localities of past volcanic action, and the relation of these phases and re- sults of volcanic action to the marked differences in our soils. The questions that present themselves to us are the following: — Have the modes of lava — disintegration, that are going on to-day in the locality of the active crater, operated in past times, and in other parts of these islands? If so, how, and to what extent, have the soils derived from the lavas been affected by those intense physical and chemical processes of rock disintegration? We shall now try to see what can be known along the line of these questions. 22 Evidences of Chemical Action In the Disintegration of the Older Lavas, and in the Formation of Soils. — The study of physical and chemical action as a factor in the disintegration of lavas, and in determining- the character of soils, has, so far, been confined to the phenomena at- tending the decomposition of lavas seen to-day at the Kilauea crater. To ascertain wliether these ])hysical and chemical causes have operated on a grand scale, and over wide areas in the other volcanic regions we have to go out and examine the lavas and soils in the several parts of all the islands. In order to closely con- nect further investigation with the observations made at the active volcano, we shall continue with the — Island of Hawaii — This island, which is the largest in the group, comprises some 4,210 s<|uare miles of sur- face. Its formation has resulted from the action of four grand centers of eruption, or "craters of the first class," viz — Manna Loa, Mauna Kea, Hualalai, and tlie crater of the Kohala mountain system, whose location is in- definite. The present active crater of Kilauea cannot be included iu this class, since it has not borne any similar part in the work of construction of the island. From a point of elevation iu tlie Kohala mountains, from which the writer made these topographic observa- tions, the system of grand craters is completely under view: with ]Mauna Kea to the left, and Hualalai to the right, Mauna Loa closes n\} tlie view, at a distance of some sixty miles, to the south. From tliis point of eleva- tion it is indicated how each of the four great mountains had an individual oi'igin and growth, first coming into visible existence above the surface of the ocean, and building up by the material of subsequent eruptions, until the huge cones were raised res]>e(t ively 1 :>,(>","> feet; 23 13,805 feef; 8,275 feet, and 5,505 feet above the level of the sea. These cones are at distances of some thirty miles the one from the next nearest one. The spaces between the cones, which were filled by the ocean during the opening history of the construction, now form the in- terior valleys, the ocean divisions not only having become closed up, but the spaces have been raised to grand plateaus, by the infilling of the great discharges from the crater cones, and lie at levels of some 3,000 feet above the sea. The filling in of the ocean spaces separating the four great mountains, and the forming of the valleys, were not wholly done by the discharges from the sum- mits of those great cones: This work was largely effected by side-flows or outbursts from the sides of the great mountains at different altitudes. Those "outbursts" be- came more or less continuous at the places of origin, and their locations are now marked by "craters of the secondary class," known frequently as "lateral cones," also as "tufa cones," wliich vary in dimensions from small mounds, that are hardly longer distinguishable, to basin — formed craters enclosing areas of many acres. In the course of the ride across the Waimea plain, and from the vantage point on the slopes of the Kohala moun- tains, the writer became able to somewhat grasp the magnificent vastness of the operations that liad gone to the building of the island! Not only had each of the four grand cones raised their burning heads out of the sea, filling space and air with clouds of steam and con- fusion, an idea of which was given to us through the recitals of the "side flow" from Manna Loa in 1868, when the stream of lava poured down into the sea, heating the water to more than the hand could bear, for more than a mile out in the ocean, and killing all living things 24 within it tliat liad uot escaped to pelagic depths, but the distribution of lateral cones bad added a still further vastnoss to the operations, and an increased confusion, and probably splendour, to the scene! In the space of a day's ride we noted no less than eighty-nine of these lateral cones, and observed mounds covered by forest on the mountain slopes which suggested many more. Each one of those, in its day and measure, was a center of eruptive force. Lava, in some form, came from their throats, and fumes and steam escaped from fissures in the lavas in their locations, leaving marks of their action in disintegration to which we must specially recur. This distribution of centers of eruptive action gives to the valleys dividing the great cones the true character of vast volcanic plains, which afford the most impressive sight of its kind that the writer has ever beheld. PEIMARY EFFECTS OF CHEMICAL ACTION ON LAVAS. In the previous paragraphs the centers of eruptive action have been divided into "craters of the first class," and "lateral craters," which we shall, for our present pur- pose, speak of as "tufa cones." The great craters, during the periods of the most vigorous activity, have dis- charged the lavas which have built up the great rock masses forming the structure of tliese islands. These rock masses indicate that the lavas were put forth at high temperatures; that they flowed smoothly, and cooled down into their present state without much change in composition. There are indications, however, that the character of the discharges from the great craters was various: Instead of tlu' high t(Mn])eratur(\ aud fnM^-fldW- < \ 25 ing lavas, lavas of lower temperature, and puddled into a state of mud by excess of fresh water, were also dis^ charged from the great craters. But these lavas do not necessarily differ in chemical composition from other solid lavas. Again, the great craters, and apparently during the period when they were approaching extinc- tion, have emitted materials — tufa, scoriae, ashes— which are more distinctly representative of the ejections of tufa cones. The lateral or "tufa cones," in the earlier period of their activity, have also produced lavas, and built up their foundations with compact and solid materials, re^ sembling the sound lavas emitted by the great craters. The material generally put forth, however, is of a dif- ferent character, whose mass is made up of fragments of altered lava, in which are found enfolded lumps of less altered lava, which are easily distinguished from the brilliant red and yellow, or brown colors of the mass by their dark or blueish gray, which is the color of the normal basaltic lavas. This material in mass is known under the name of "tuff," or "tufa." Dana defines tufa "as a rock not very hard, made from comminuted volcanic rock, more or less altered by the action of steam and vapors." He continues "the tufa made from those igneous rocks that contain iron-bearing minerals, such as basalt, is usually yellowish brown or red in color. ' Lyell defines tufa as "composed of small angular fragments of scoriae, and the dust of the same, produced by volcanic explosions." Leonhard also speaks of tufa and says "volcanoes produce, in addition to the solid lavas formed from the flowing lava streams, j)eculiar masses of ejected materials which are found in less or greater proportions around the borders anical s])eci- mens of lavas in the iuniiediate localitv where samples 27 of soil were taken. Many of tlie specimens of solid lavas, and of tufas, have been examined, and in the following tables are found strictly typical representatives of the two classes of rocks. The analyses, by reason of stress of work, are only partial, and are confined to observing the behaviour of the elements that would indicate the change of chemical composition on a grand scale in the passing over of the fragments, from solid lava masses, into tufa under the action of acid vapors. The first tables are given to solid lavas, dividing these into non-hi/droiis and hjidrous lavas, and the last table to tnf(ii<. NON HYDROUS SOLID LAVAS. Moisture. Combined c, o Water. ''^ '^'^ FeO Fez O3 AI2 O3 Ca Per cent. 0.73 0.04 0.25 0.49 0.63 0.36 Per cent. 0.28 0.09 31 0.00 08 07 Per cent. 48.18 52.51 48.84 52.89 47.85 44.45 Per cent. 8.40 9.94 9.69 6.51 7.66 8.39 Per cent. 4.42 3.33 2.42 3.64 4.20 6.23 Per cent 18.02 23.71 23.17 17.80 28.62 20.32 s ID Means 0.41 13 49.12 8.43 4.04 21.94 9.24 HYDROUS SOLID LAVAS. Moisture. Combined Water. Si02 Fe FezOs AI2O3 Ca Per cent. Per cent. Per cent. Per cent. Per cent Per cent. 0.33 0.90 42.19 6.71 7.94 12 07 1.95 1.83 47.42 8.33 4.18 14.30 s 1.05 2.53 44.85 9.91 4.17 20.44 V 0.81 1.00 46.29 8.87 3.82 14,30 0.10 4.27 40.95 11.17 5.73 19.74 0.70 1.34 46.50 7.00 6.87 21.51 Means 0.82 1.96 44.66 8.66 5.62 17.60 8.23 28 TUFAS. Moisture. Combined Water. SiOz FeO Fe2 O3 AI2 O3 Ca Per cent. Per cent. Per cent. Per cent. Pel cent. Per cent. 3.97 12.98 15.92 0.09 49.02 13.42 7.62 9.86 25.70 1.23 25.07 26.26 10.63 11.15 23.72 1.84 16.34 32.56 8 9.52 11.16 29.46 2.93 19.17 22.53 7.25 11.23 31.69 2 67 20.26 20.66 9.44 9.27 39 42 0.79 15.78 20.18 7.16 11.37 25 42 1.66 26.77 20.23 6.22 14.93 21.52 2.37 33.90 32.87 Means 7.72 11.49 26.60 1.69 25.79 23.58 1.41 The analyses given could be added to; but we have con- fined ourselves to analyses made by our laboratory, the specimens analyzed having been collected hj the writer as typical. The analyses recorded are also selected in order to con^ vey a view of the extremes of variation, which in the tufas are very great. Before discussing these variations we shall bring the general results together in a table of averages. Lavas. Moisture. Combined Water. Si02 Fe Fe2 O3 Ah O3 CaO Non— Hydrous . . . Hvdrous Tufas Per cent. 0.41 0.82 7.72 Per cent. 0.13 1.96 11.49 Per cent. 49.12 44.66 26.60 Per cent. 8.43 8.66 1.69 Per cent 4.04 5.62 25.79 Per cent. 21.94 17.06 23.58 Per cent 9.24 8.23 1.41 As distinguished from the less crystalline, compact non-hydrous lavas, the hydrous lavas not infreciuently show separations from the lava mass of silica, also of lime, which occupy the larger vesicles, and indicate that an initial disintegration of the solid lavas has taken place. 29 The analyses given of tufas convey to us an idea of the different degrees of action of the steam and acid vapors to which the blocks and fragments of solid lava were subjected in passing over into the tufa mass. In one specimen nearly 40% of silica is still found, and in an> other the silica is reduced to merely 15%. These very significant variations in the silica and other elements we shall recur to in more detail under the heading of Chemical Action on Lavas After Emission. Other effects, showing how intense the chemical action has been at the times of ejection of tufa lava materials, are seen in the almost complete oxydatiou of the ferrous iron, and in the removal of the lime; results that are most instructive in the study of rock decomposition un- der the action of various causes. The main conclusion that is drawn for us by the analyses, and which is set forth briefly in the table of averages, may be expressed thus: Amongst Haicaiian lavas are those which have been discharged from craters^ -flowing and cooling into rocks having the composition of normal basalts. Others, origin all if of the same composition, have undergone such alteration that they noio compose rock masses having a radically different chemical composition and color appearance. This alteration took place at the time of ejection, and under the action of chemical causes, and previous to the later action of secondary causes of rock disintegration, such as "'Weathering,^- which has apparently been the only agent of disintegration of certain of the normal lavas. The effects of these primary chemical causes of disin- tegration which are exhibited in the appearance of the tufa masses, and marked by every variation of colors be tween the most vivid reds and yellows, and upon which the marks of fire are still as clear as upon bricks fresh 30 from tlic kiln, will be fnrtlier coiisidonMl at a later time, aud ill their relation 1<» the soils derived from the tufas. At present we shall consider other chemical causes that may have operated in determiniuj>- the character of soils. CiiEMic^u. Action on Lavas after Emission. — In previous paragraphs we have considered the operation of steam and acid vapors that is going on at this time u])on the lavas over large areas on the floor of the active crater at Kilauea, and over the areas surrounding the crater. We have shown that the solid lavas are undergoing rapid disintegration under the chemical agents that are operat- ing, and are being resolved into the decomposition pro« ducts that have been set forth. The present purpose is to try to find out if these chemical causes, that are operating today at Kilauea, have operated in other localities, and over extensive areas, during previous periods of time. In the general review of the to])ography of the island, we spoke of the great craters, and of the general distri- bution of tufa cones, as seen from the Waimea plains. We must noAV come down, and with greater precision and detail, to the examination of the districts where, to-day, we are growing cane, and from which we have analyzed abundant samples of soil for our guidance in fertilization. In the first place, it was necessary to note whether great volcanic action has transpired during previous periods of time in the localities that are under considera- tion, as indicated by remains of extinct tufa cones, or whether all the lavas that have formed the soils in those localities have flowed from the great craters that are many miles away? Commencing in the district of Kan, on the final slopes of Manna Loa, around and between Honuapo and Pahala, several fine specimens of tufa cones are located. In the immediate neighborhood of these 31 cones the rock masses are marked by the strougest signs* of alteration, and vast layers and deposits of red, yellow, and whitish earths are found. Leaving Kau, and pass^ ing over by way of the Kilauea crater and down into the district of Hilo, on the other side of Manna Loa, we come upon further centers of past volcanic action, and at levels close by, or up to 1,000 feet above the sea. One crater on the Kau side was noted where the bottom had dropped and let one-half of the cone down into the sea, thus revealing the grand throat of the crater, and the lurid colors of its interior, w^hich varied from white, to yellow, to red, to black. At Hilo there is the cluster of cones known locally as the "Halai Hills," three in num- ber, and of moderate dimensions. Cane is now being planted at their feet. Above Wainaku are several small cones, and a series of profound caves, to which the writer's attention was called by Mr. John Scott. At Onomea, which has been a center of volcanic action of the most violent order, as shown by the great gulchea and the irregular surfaces of the land, there is a further cluster of tufa cones that form a marked feature of the district. As we must return after this cursory descrip^ tion to discuss the special marks and characteristics of each of certain centers of past volcanic activity, we shall, for the present, merely state that in the course of repeated rides between the town of Hilo and the grand gulch of Waipao, near Kukuihaele, the writer noted thirty-nine extinct craters, varying in size from areas of an acre, to enclosed basins containing several acres. During this ride of some sixty miles, more than fifty gulches were crossed, some of them of the first class in depth and grandeur. The thirty-nine craters noted are all upon levels at which cane is being planted, the basins of 32 some of tliom actually bearing cane, and otlici- jLivowths. Alono the same coast of Hawaii, in the district of Ko- hala, are also fonnd several lateral cones. It is thus seen that throuji'hoiit the districts reaching- from Ililo to Ko- liala there are areas that are covered by the centers and indications of past volcanic action, the evidences and results of wliich we liave now to consider more minutely. We have, in a case of specimens before us, small blocks of partially decomposed lava that were taken by the writer from lavas in place which form the irrej>ular walls of several jiulches. One collection of these blocks waa from a oiildi running through the sugar plantation abctve AVainaku; a second from Kawainui gulch near Onomea; and a third collection from a gulch below Kukaiau. In each case, these blocks were obtained by removing several feet of the front of the walls, and getting into small in terior caves which were formed and left by past volcanic- movements. Some of these caves are amply large to move freely in, and they consist usually of an arched roof over a more or less level floor. The floors of the caves from which the three collections of partially decomposed lava were taken, Avere covered with a (h'posit, almost white in color, and from one-eighth to one-fourth of an inch in thickness. The greatest care was taken to ascertain that this white material was brought up from the lava ui)on which it lies, and not drained from the lava of the arch above by the action of water. Examination of tlie lava under the deposit indicated that the material had been brought up by ascending steam. Bh)cks or pieces of the lava forming the floor of tlie caves were taken out, packed Avith cai(% and brought to our laboratory, and are pre- served, Avith Ihe deposit of Avliite material fresh and undisturbed upon them. The freshness of the deposit is 33 remarkable; especially upon the specimens taken from the cave in the gulch above Wainaku. In the laboratory, this deposit is found to consist, in the first place, in more than one-half, of silica, considerable alumina, less iron, and small quantities of lime and sulphuric acid, no car- bonic acid being present. An exact quantitative state- ment is unreliable, due to the crumbly state of the lava, from which the dex)osit cannot be removed without in- cluding portions of the lava. In the search for such caves as we have described, and for deposits of materials that were possibly to be found in them, the writer was endeavoring to locate centers of past volcanic activity where chemical action upon emit- ted lavas may appear to have produced results corres- ponding to the effects observed to-day at Kilauea, where the action of acid vapors is seen to be resolving the lavas into the products described in previous paragraphs. The conditions of those caves that we have examined, which belong to a period of volcanic activity that prevailed in- definite ages ago, correspond to some of the conditions of location and environment marking the places of the present chemical action near the volcano, and the deposits upon the specimens of lava taken from those an- cient caves bear, in appearance, a most palpable re- semblance to siliceous deposits upon the blocks of decom- posing lava that we took fresh from a center of extreme chemical action on the borders of the Kilauea crater a year ago, the analysis of which gave 70.8% of silica; 16.1% of iron; 7.3% of alumina; 4.8% of lime, and 2.2% of sulphuric acid. In addition to the cave deposits, there are other pro- ducts of lava-disintegration, under the action of chemical causes, which appear to correspond to materials into 3 34 ^vhicli the lavas are being resolved at Kilauea, the com> position of the latter we have already given. These are layers and vast pockets of red and yellow earths, more or less consolidated, and also of earths whose colors vary from a dirty or yellowish white into actual blue. Al- though the number of caves does not exceed four in which we have been able to identify the evidences of chemical action upon the lavas with a sufficient measure of reli- ability, the distribution of the other products of lava- decomposition by chemical agents is on a grand scale. We have examined and identified the several kinds of colored earths in all the gulches around Hilo and Papai- kou. Further in the gulches named Kahalii, Kawainui, Kulaimanu, Kapaheehee, Kolekole, Hakalau, Maulua, Kapehu, Laupahoehoe, Kawali, Ookala, Kukaiau, Puu^ ohihaihai, Honokaa, Kukuihaele, and in the gulches and cuttings in the district of Kohala. The recent cuttings made in the building of roads in the districts of Hama^ kua and Kohala, likewise of Hilo, have furnished ample opportunity of noting the distribution of the earthy decomposition products under discussion. These earthy products are found in alternating red and yellow layers, of thicknesses varying from three inches to more than one foot. The layers repeat themselves, as shown by the cuttings, one series being found only two to six feet be- low the ground surface, whilst other layers are lying even to twenty feet below the surface, and probably still deeper, which is indicated by their position in the bluff formations overlooking the sea in the Hamakua district. In addition to being seen in irregular layers, the red, yellow, and whitish substances exist in vast deposits or pockets, whose masses may be expressed in millions of tons. These substances however, are not 35 found uniformly throughout all the lava formations. In what appear to have been centers of great volcanic and chemical action the formations have been largely re- solved into these separate substances. But these centers are more or less acutely defined. They cease suddenly, and we come upon areas of formations that are under^ going gradual disintegration, the dissolved materials be^ ing borne away in rain waters into the sea as they are released. So that there are areas where the lavas and formations appear to have undergone violent disintegra^ tion, and other vast areas where the evidences of such violent action are almost totally absent. Further generalizing upon the extent of the areas where it is indicated that modes of violent chemical dis- integration have operated, and upon the appearances of the several decomposition-products, will not add to our knowledge. We must ascertain the chemical composi- tion of these disintegration products, and then bring them into comparison with other similar products, the agents and modes of whose production we have identified. At the time of the field examination of those products of ditsintegration, in the districts of the four islands, we took typical samples of the several kinds of earths. Some eighty samples, in all, were taken. On account of stress of work, this number was gone over again by the writer, and reduced to forty samples, these representing localities and centers, bearing the marks of the most evident volcanic and chemical action, upon the four isl- ands. The analyses, which are found in the followina tables, are made upon the materials as they were found in place. The elements included are those whose varia- tion in the several kinds of earths bears most closely up« on the immediate question. The estimation of the sul- 26 pliuric acid in each sample is indispensable on account of the part which it is believed to have played in the pro^ cesses of rock-disintejiTation under discussion. The rel- ative significance of the presence of sulphuric and car- bonic acids will be considered in a later paragraph. In the first table we give the analyses of the decom- position-products found in localities on Hawaii, after which corresponding evidences of the same character of rock-disintegration by chemical action that has transpired upon the other islands will be briefly given. LAVA DISINTEGRATION-PEODUCTS, ISLAND OF HAWAII. Locality ■ Hilo Hilo Hilo Honomu Paauilo Ookala Ookala Lauijahoehoe. Lanpaboeboe. Laupaboeboe. Hamakua . . . . Halawa Kobala Kobala Com- 1 Color. Mois- bined SiOa FeO Fe, 0, AU O3 Ca 0| ture. water Per Per Per Per Per Per Per cent. cent. cent. cent cent. cent cent. Brown 86.82 11.70 11.53 4.27 10.45 19.42 1.25 Red 19.82 9.64 21.80 0.89 23.04 22.841 0.49 Yellow . . . 9.45 15.56 17.66 3.64 20.64 80.99 0.60 Red 17.51 4.74 25.10 0.49 31.12 18.95 0.40 Red 28.52 5.84 16.27 0.98 25.28 21.42! 47 Red 16.69 5 85 26.17 0.70 27 50 21.9110.22 Yellow . . . 4.15 27.85 2.26 0.77 2.40 61.12| 30 Li^bt gray 30.08 10.80 27.88 0.61 1.86 28.001 0.62 Red 15.25 4.51 26.60 0.84 36.00 15.14! 0.49 Wbitisb . . 2.47 30.52 2.44 1.50 1.64 59.841 0.50 Grav 7.66 5.79 37,40 5.74 12.24 18.801 0.30 Wbitisb 18.26 11.74! 34.04 1.26 4.40 84.68 0.25 Wbitisb . . 5.16 25.881 10.96 0.77 7.36 48.14| 0.45 Reddisb . . 9.54 1.03 38.37 1.42 16.00 25.92 1 2.19' SOa Per cent. 0.67 0.80 0.38 trace 0.07 trace 0.50 uoue 0.40 0.88 0.48 0.14 0.35 1.50 CO2 Per cent. none 0.10 0.25 none none none none 0.10 0.25 0.20 none none none none These analyses were made precisely the same as those of the lavas, and thus the data express the amount of the elements in the total substance. As the Island of Hawaii, excepting probably the Ko- hala division, is the most recent in the group of islands, and its products the freshest, we shall call attention to the table of data before proceeding further. Without noteworthy exceptions, it is seen that Avhere the "combin- ed water" is high, the proportion of alumina is also high, 37 and in instances enormous! The earths having these high contents of alumina and combined water are a dirty white, gray, or yellow in color. When glowed, for the removal of the combined water, the color may remain yellow, or becomes pink, or goes over instantly into a vivid red, which changes are determined by the propor^ tion of iron and the form in which it is present. The yel- low hydrate of iron at once becomes red on glowing, bnt the silicate changes slowly, and very little. Very striking is the variation in the contents of silica, and of iron, in the substances. In examples we find the silica as low as 2%, and the iron as high as 37%; yet im stances are found where the silica and iron are both reduced to almost nothing, and the alumina forming ah most the total of the earth. It has also to be specially understood and kept in mind that these substances which are so totally different in their composition, are found in the closest contiguity to each other. The Ookala "red" and "yellow," although each represents thousands of tons of material, are lying near to each other, yet separated by acute lines of division. It is also, and even more strikingly, the case with the "red" and whitish earths taken in the locality of Laupahoehoe. The writer, in some cases, has taken samples of "bright red," and of "whitish yellow/' from heaps of these earths not more than three feet apart. This close contiguity of earths of such totally different chemical compositions indicates most clearly that a cause other than "weathering" has operated in the rock-disintegration. Island of Maui. — This island is said to be older than Hawaii. To the writer, the indications of age resemble those of the Kohala division of Hawaii, Three of the analvses are of earths from the region of 38 greatest volcanic activity, folloAviiig the line of lateral cones that descend the slopes of the grand crater of Hale- akala to the sea. The writer located some six of these lateral or tufa cones. Two more are of products separate ing out in the disintegration of the lavas of West ]\[aui, which was separated formerly from East Maui by a channel, in whose place now lies the isthmus that tiea the two great crater mountains together. The borders of West Maui are skirted in localities by coral formations, which formations underlie some of the later lava flows. It is also indicated, as upon the Island of Oahu, that later volcanic activities have disturbed the underlying coral, or limestone, whereupon the escape of steam and vapors from below took place, bringing up vast quantities of carbonate of lime, of silica, with small amounts of iron and alumina. The deposits of these materials are a feature of the mountain slopes of West Maui running down to the bay of Lahaina, and the writei* has frequently been asked by people passing the island on steamers "what the deposits are?" The districts be- ing small, the number of analyses has been reduced to a minimum; yet numerous samples of earths were taken corresponding to these analyzed. LAVA DISINTEGRATION-PEODUCTS, ISLAND OF MAUI. Locality. Color. Paia . . Makawao . . Makawao . . Olowalu . . . Lahaina red red yellow white white 1 Mois- ture. I Com- 1 billed' water etc. Si O2 Fe O Fe^Oa AI2O3 Ca O S O3 C O, Per Per Per Per Per 1 Per Per Fer cent. cent, cent cent. cent. cent. cent, cent 10.S5i 18.83 29.90 0.22 23.10 26 00 0.31 12.601 13.39 5.66 3.92 58.00 .5.06! 0.61 30.02.32.63 7.66 O.OO] 1.50 29.06 1 0.30 3.44 13.85 10 80 trace trace 3.06 37.36 1.66 11.51 12.18 O.8OI 4.32 5.28| 40.24 Per cent. 0.31 none 0.03 none 0.02 none 0.05 32.66 0.10 30.90 39 The "red" and "yellow" earths from Makawao were taken by the writer from places only three yards apart; but they represent bulks of thousands of tons. The crater region of Makawao abounds with these different kinds of earths, in which the guava tree is flourishing; and this tree is preparing the earths for growths of a more delicate nature. The white deposits in the districts of Olow^alu and La^ haina are not to be regarded as lava disintegration-pro^ ducts. Their prevalence, however, indicates to what an extent the areas of lava flows have been marked, during a past period, by escaping steam and vapors emanating from the depths below\ Island of Oahu. — This island being the chief island of residence, its volcanoes are better known than others. Dana says the island is derived from the operations of two craters of the "first class." We are able, however, to locate not less than twenty lateral or tufa cones, cer- tain of Avhich are of notable dimensions. Better than upon even the younger islands can be observed the marks and results of escaping steam and vapors upon the lavas after emission from the craters. The slopes of Diamond Head and of Punchbowl are white with deposits of silica and carbonate of lime, and the fractures of the lava masses, and large fissures, are filled up with these materials that have been left behind by the escaping vapors. As on West Maui, the presence of more or less of car- bonate of lime in the deposits is due to limestone or coral formations upon which the latest eruptions overflowed. Even the lavas of these districts contain nearly 15% of lime, indicating that the limestone below was melted and mixed up in the up-coming magma, within which it is 40 now seen in white clusters of beautifully ciystallized carbonate Avitli some silica. The following analyses are of disintegration-products of lavas from localitites where the coral has not ent(»red into their composition. DECOMPOSITION-PRODUCTS OF LAVA, ISLAND OF OAHU. Locality, Color, Tantalus reddish . Tantalus ... . ibrown . . Tantalus .... brown .. Kokoloea . . . brownish Kokoloea . . . dark .... Ewa .. white . . . Ewa whitish . . Waialua i reddish.. Heeia crimson . Mois- Com- bined Si02 FeO FezOa AI2O3 t CaO SO3 water Pe- Per Per Pe Per Per Per Per cent. cent. cent. cent. cent. cent cent cent 15.10 10.74 29.14 1.60 15.20 26.10 0.40; 0.10 1.27 30.00 6.70 0.44 8.45 52.10 0.43 0.20 fi.37 21.04 12.84 1.99 13.29 43.90 0.53 0.35 12.11 9 57 33.96 0.75 11.14 30.14 0.40 0.20 1.85 1089 7.52 1.40 70.40 6.30 0.62 0.56 1.60 4.78 87.89 0.69 1.85 2..33 0.20' 0.07 3.84 9.80 55.96 0.00 9.92 19.48 0.77 0.15 7.91 1.15 29.29 1.80 26.06 22.07 1.88; 1.53 5.21 6.84 83.70 2.56 18.66 23.71 1.481 1.151 Per cent. none none none none none 0.46 none none none These data repeat the indications set forth in the ex- amination of the substances from Maui and Hawaii. The Kokoloea substances were taken from places separated only a few yards from each other, yet the one is largely an alumina compound, whilst the other contains over 70% of iron oxide, of which Ave shall speak later. Special- ly noteworthy are the high amounts of silica found in the Ewa products that were taken by the writer from two deposits that are still in a. very fresh condition. The large amounts of sulphuric acid found in certniii of th(* products furnish a clear indication that the steam vapors that acted in the disintegration of the lavas were highly sulphurous. IsL.sjs^D OF Kauat. — Tlio geologists appear to agree in considering Kauai the oldest island in the group. We have reason to consider, however, that whilst Kaiiiii mav 41 have appeared above tlie ocean before any of the islands, volcanic activity continued upon it after the other islands came into existence. But the greater age of the forma- tions and lavas is indicated by given disintegration-pro^ ducts notwithstanding. Geologists cannot locate with certainty the location of the primary craters which laid the foundations, and built up the main part of the structure of the island. The centers of lateral crater action, however, are almost more definitely marked than upon the younger islands. At the north end of the island in the district of Kilauea, the lavas, and the soils derived from them, indicate a j)ast period of extreme lateral activity. Crater hill, one part of which appears to have dropped down below the sea, leaving a bluff wall which had been the interior of the throat of the crater, is, with other tufa cones, an abiding evidence of past volcanic action. In the district of Kealia the grand gulches and tufa cones repeat the same state^ ment. The Kilohana crater gave forth much of the material that has formed the soils in the district of Liliue, although all Lihue soils do not appear to have been deriv- ed from one source. From Lihue we pass on to Koloa, where we come upon a region of past volcanic activity that is almost without a parallel. Within an area of a few thousand acres are seen what have been nine, and possibly eleven, crater cones; one line of which — the "Button craters" — corresponds to the Diamond Head series on Oahu, and to the Makawao chain of cones on Maui. Analyses have been made of only seven products from Kauai; but these are so extremely distinctive in character that further analyses could hardly add to the indicationa that they are calculated to convev. 42 DECOMPOSITION-PRODUCTS OF LAVA, ISLAND OF KAUAL Locality. Color, Mois- ture. Per cent. Koloa White. .. 4 44 Kolo;i ; Yellow... 10.13 Kolo.i Yellowish 24.88 Koloa White ... 830 Lihue Red 4.37 Kealia Red 2.24 Kilauea Red 1 5.34 Com- bined SiOa water Per Per cent cent. 15.55 2 31 10.41 32.52 5.87 43.45 12.47 42.42 10.23 9.10 10.41 12.98 8.90 4.96 FeO Per cent 0.00 0.94 0.00 0.00 0.84 1.92 1.93 Fe2 03 AUO3 CaO Per Per Per cent. cent. cent. 0.85 1 09 31.24 11.37 2417 3.59 3.84 18.07 0.46 2.24 33.76 0.25 63.68 9.72 0.66 62.72 8.22 0.50 67.06 11.19 0.32 S03 C02 Per I cent. I 44.65 I 4.65 trace 0.10 0.54 [trace 0.59 Per cent. none 0.14 none none none none none We did not succeed in finding distinct siliceous deposits on the Island of Kauai. The composition of the other products, however, indicate the enormous amount of silica that has been released by the disiutetiratiou of the lava, and removed elsewhere. The undecomposed, vesicular lavas in the region of Koloa and Lihue are quite remarkable for the amounts of pure silica contained in the vesicles, and deposited on the surfaces of the lava. In Koloa we have an example in process of JmoUnizatlon, and on a large scale. The writer has in his collection a specimen of one hundred pounds in weight, obtained with the assistance of Mr. Anton Cropp, in which is beautifully illustrated the mode of coming together of the alumina and silica, and the production of an almost theoretically pure kaolin, and the separation of the other elements of the decomposing lava. Upon the plain of Mahaulepu, in the Koloa district, we found the most extraordinary piece of evidence of pre- vious volcanic action, and of the mode of the subsequent disintegration of the lavas. Side by side, and on areas extending over a considerable portion of the plain, are found deposits of red and yellow cjirths. Tlie yellow eai'ih coiitjiins no less tlinn 5.82%, upon its dry Aveight, 43 of sulphuric acid, aud sulphate of lime is found in crys- talline form, aud amounts to over 9% of the earth, from which it is separatino- out almost chemically pure. In the district of Lihue, Kealia and Kilauea, and in what have been centers of extreme volcanic activity, the separation of the elements which formed the lavas has proceeded to more ultimate lengths. The layers of red earth, some of which are found in deep cuttings, and twenty to thirty feet below the land surface, and others only four feet below the surface, have consolidated into hard concretions of almost pure iron ore, or hematite. These concretions we have only succeeded in finding on the two older islands — Oahu and Kauai — and they, with the kaolinization proceeding at Koloa, indicate the con> tinuance of lava-disintegration, and separation of the products, under the agency of "simple weathering," and after the chemical causes, of which we are speaking, had ceased. It is understood that whatever may have been the character and extent of the primary chemical action upon lavas, the disintegration has been continued and completed by the several climatic influences that are ex- pressed by the term — "weathering." Having examined the products from each of the four islands which may be considered as evidence of the opera- tion of chemical causes in the disintegration of lavas in certain localities, we shall now bring these ancient pro- ducts into comparison with the substances that are being produced today at Kilauea, on the borders of the active crater, by the action of sulphurous steam on the lavas. We may here add that, in addition to the samples repre- sented by the analyses recorded, we have made quali- tative examinations of the other almost chemically pure samples of gypsum, siliceous, and other compounds in 44 order to justify any allusion to such substances that it did not appear necessary to examine quantitatively. In the following comparisons of recent and aiicieuf products of disint(\iiration the data give the relative amounts of the elements free from water and volatile matters. Thifo is necessary on account of the extreme variation in the proportions of volatile matters in the several substances, and in order to further compare the composition of the several products Avith that of the original lava, which is almost free from volatile bodies. YELLOWISH-VVHITE-OOLORED, SILICEOUS PRODUCTS. Age of Products. Si O2 FeO FeaOa AlaOa Ca SO3 Recent— Kilauea Cra^^er Per cent. 88.10 Per cent. 0.90 0.73 Per cent 4.00 Per cent. 1.51 Per cent. trace 0.21 Per cent. 1.33 Ancient— Extinct Crater Regions 93.88 1.97 2.49 0.08 GRAY-COLORED, SILICEOUS PRODUCTS. Age of Products. sio. FeO Per cent. 2.23 0.00 Fe,0., Per cent. 9 65 11.49 Al„ 0, Ca 0., Recent- Kilauea Crater Ancient— Extinct Crater Regions Per cent 67.69 63.45 Per Per cent cent. 8 38 4 45 22.57 0.89 Per cent. 2 12 0.27 In the following comparison, lime carbonates are ex^ eluded from the average of "ancient" products. As we have already explained, these carbonates are derived from lime formations underlying lava flows, and are not true disintegration-products of lava. 45 WHITE, CALCAREOUS PRODUCTS (GYPSUM). Age of Products. Recent — Kilauea Crater Ancient— Extinct Crater Regions SiO, Per cent. FeO Per cent. 0.00 0.00 Fe.O. Per cent. 0.70 1.06 AUG., CaO Per cent. trace 143.40 Per cent SO. Per cent. 44.73 1,36 39.04:55.81 It is difficult to make a comparison, of any value, be- tween the alum products that are being formed from the decomposing lava at the Kilauea crater with ancient deposits in which alumina predominates. The Kilauea alum deposits contain, in addition to iron sulphates, con^ siderable sulphates of magnesium and of the alkalies. In the ancient products these sulphates have been almost washed away, which has resulted in increasing the amounts of alumina and silica, GRAY, ALUMINA PRODUCTS. Age of Products. SiOa Per cent. 0,80 8.97 FeO FezOs CaO AI2O3 SO., and Ha Recent — Kilauea Crater Ancient — Extinct Crater Regions Per cent. 0.00 1.44 Per cent. 12.30 7.80 Per cent. 0.50 0.34 Per cent. 26 2 45.5 Per cent, 45.65 25.60 As already remarked, these ancient alumina jjroducts do not admit of close comparison. They also unquestion^ ably owe their present composition to slow processes of change that followed the primary chemical action upon the lavas. Moreover, it must be borne in mind that alumina products, resembling these under discussion have resulted from the decomposition of rocks in older countries where it has not yet been shown that chemical causes of disintegration have played any part. In the following and last table we brino- into com- 46 pari son the red-earth products obtained fresli from the active crater where they are now forming, and tlie red earths collected from all parts of the four islands. Owing to the great variation in the ancient red products, which variation appears to relate to the variation in age, we give first, the composition of the recent red product from Kilauea. Second, the composition of an ancient red earth that corresponds most closely with the recent pro- duct. Third, the ancient red product that contains the least amount of iron oxide. Fourth, the ancient red pro- duct that contains the largest amount of iron oxide. Lastly the mean composition of all the red earths. This division is made in order that the variation in the compo- sition of these red earths shall be understood, and for the further reason, that these earths are a factor of im- portance in the composition of some soils. EED, IRON PRODUCTS. Age and State of Products. Recent- -Kilauea Crater Product Ancient — Corresponding Product , Ancient Minimum Iron oxide content . Ancient— Maximum Iron oxide content Ancient— Mean of Red Products SiO„ Per cent. Fe.Os AloO., Per Per cent. cent. CaO Per cent, 32.50 44 5018.10 0.20 33.15 44 8618.87 60 42 9019.48 28.98 7.65 83 68 6.84 28.7147.32 21.80 SO, Per cent. 1.57 0.50 2.45 1.68 0.82 0.04 0.90] 0.60 The examination of the products of lava-decomposition of an earlier period, and the comj)arison of those with the products of disintegration that are being formed, at the present time, in the region of the active crater at Kilauea, have led to very significant results. The location of the ancient products in what have been centers of past, ex- treme volcanic activity; the marked resemblance of these products, in appearance and chemical composition, to the substances now in the act of formation from disintegrate 47 ing lavas, by visible processes, — all these phenomena in- dicate that the "recent" and the "ancient" products, under discussion, have had a similar origin. Moreover, these phenomena indicate that in the disintegration of the older lavas, chemical causes have exercised a strong action, over extensive areas, and j)revious to the later action of "simple weathering," the results of which will be found in the present character of the soils derived from them. LATERITES: THEIE OCCUREENCE AND ORIGIN. In the previous paragraphs we have endeavored to compare the products of lava-disintegration of a i)ast period with the substances into which the same kinds of lavas are being resolved to-day, but we have not, so far, attempted to definitely designate those products. It is evident, however, that in the highly siliceous, light-yellow colored, products we have substances that will ultimately consolidate into siliceous sand-stones, ex- amples of which are already found in a very developed crystalline state. Further the calcareous products have, and are still resolving themselves into definite deposits of almost pure gypsum and carbonate of lime; whilst the alumina is separating as masses having a remarkable degree of purity, exceeding that of the hauxltes of France and Ireland. Each of these products has a definite char^ acter, and they all differ radically from the "red earths;" which, from their appearance, physical character, and chemical composition may come under the designation of laterites. Sir Charles Lyell describes laterite as a "red or brick- like rock, composed of silicate of alumina and oxide of 48 iron;" wliicli description exactly agrees with the cheiuical composition of tlie red product that we have found to be forming, at the present time, near the active crater, and of the red earths collected from centers of past volcanic activity. Concerning the occurrence and geographic distribu- tion of the laterites, Lyell says further, "The red layers, called 'ochre beds,' dividing the lavas of the Giants Cause- way (England) and the Inner Hebrides, appear to be analogous to the laterites of India, Avhich were found by Delesse to be basalt (trap) impregnated with red oxide of iron, and in part reduced to kaolin." Blandford in his "Geology of India," speaks of "high level laterite as non- detrital, or iron clay;" which Posewitz says bears a closer resemblance to the laterites examined by him in Banga than do those which occur in given regions in Africa, where they were examined by Peschuel-Losche, and are described by Sachsse in his "Lehrbuch der Agricultur- chemie." Wohltmann speaks of a laterite formation from volcanic, basalt rocks in Liberia. Credner treats of the laterites of South America, and compares these with the laterites of India. These substances, which are better knoAvn as "ochres," are found distributed over most sec- tions of our globe, and independently of the present climatic conditions; although, at this place, it must be borne in mind that countries which are distinguished to- day by temperate^ climates, during a previous period exhibited the conditions of tropic lands. In the British Islands, over Europe, down to the Mediterranean, the "ochres" are found, and utilized in the arts and manu- factures. Over North America vast deposits of "red ochres" are widely known, and which, a manufacturer of mineral paints in a city in the United States says. 49 ^'are the same thing as the red ochres sent to us from the Hawaiian Islands, for which we have no use, as we have too much near home." At present, we are not able to judge to what extent it may ultimately be found the rule, but the occurrence of the laterites, has been largely found in connection with igneous rocks. Lyell not only says that the "red ochres" in England "divide the lavas;" but continues "we feel sure that the rock of Staff a, and that of the Giants Cause- way (England), called basalt, is volcanic, because it agrees in chemical composition with streams of lava known to have flowed from craters."- Lyell says further "these basaltic and other igneous rocks are associated with beds of tufa in various parts of the British Isles;" and "the absence of cones and craters in England may be due to the eruptions having been submarine." In Madeira and the Canary Islands, Lyell again says the lava flows are divided by red layers of laterite. Blanford says the "high-level" laterites of Central and Western India are found lying upon trap rocks (basalt), that are of igneous origin. In this matter, however, our knowledge is not exact and universal enough to speak emphatically on a relation of the laterites to rocks of volcanic origin. Moreover, it appears from the observations of other in- vestigators that laterites occur where it does not yet appear that igneous rocks are found. Bearing upon the question of the orU/lii of the laterites, Lyell says "the red bands or layers of laterite are prob- ably ancient soils formed by the decomposition of the surfaces of lava currents. These red soils may have been colored red in the atmosphere, or burnt into red brick by the overflowing of heated lavas." Posewitz considers the formation of laterites is due to the superficial weath- 50 ^ring of rocks and soils. Woliltmaim says "the mode by which the laterites have been formed is not at all under- stood. Some have erroneously taken these products to be a result of the action of sea water on rocks, or sedimentary deposits from sweet water; whilst others have as erroneously considered them in some ^v£iy pos- sibly connected with volcanic movements." Again he says "it must be regarded as a geological feature of latO'- rite, that it has only been formed where the processes of weathering and leaching have gone on for thousands of years. The expiration of numberless thousands of years of the action of weathering upon the materials of the earths' crust Avas an absolute necessity in the forma- tion of the laterites." Concerning the origin of the red ochres or laterites of the Hawaiian Islands we do not need to go into a further lengthy discussion. The minute description of the disin- tegration of lavas in the region of the active crater at Kilauea, that is proceeding at this time, whereby the solid lava is being resolved, by the action of sulphurous steam, into the several products of decomposition, of which a red earth, or laterite, is a prominent one, does not need to be repeated. Also our account of the distri- bution of centers of past volcanic action over large areas, and the comparison of the "ancient" products of lava disintegration with the "recent" products that can be seen and obtained to-day — all these things place before us indications and proofs, of the most undoubted charac- ter, that the formation of laterites upon the Hawaiian Islands has been, and is still being, due to the intense action of acid vapors upon lavas, whereby these are being resolved into the siliceous, calcareous and earthy bodies which we have already fully described. 51 In evidence of the mode by which these "red earths" have been formed in past periods, it was in the first degree important to bear in mind the agents by whose action the formation could have been carried on. For that reason, in the examination of all the products of lava-disintegration the sulphuric and carbonic acids were invariably determined. If the formation of the laterite had been due to the superficial action of weathering upon lavas and soils, the more active agent in the process must have been carbonic acid, derived from vegetable decay, and evidence of its presence and action might be deduced from iron carbonates found in the products. If, on the other hand, the rock-disintegration had been primarily caused by the action of acid vapors escaping from below, and sulphurous or sulphuric acid had been, as it is found to-day, the acid element in the steam and vapors, then it was expected that sulphuric acid might be found in the laterites and other disintegration-products, and especially in such as were found at great depths from the surface, and which rains had not leached. If we recur to the analyses of the "red earths" found in localities upon the four islands we see that in most of the products no carbonic acid, but very notable quanti- ties of sulphuric acid were present. To save the reader the necessity of looking back over the several tables we reproduce the sulphuric acid contents of the recent "red earth" (laterite) forming at the present time at Ki- lauea, and of certain of the laterites formed during earlier periods of volcanic action at centres over large areas of all the islands, to which are added the amounts of the acid also contained by certain of the kaolin products, and the highly siliceous separation-products, commonly call- 52 ed "yellow ochres," and also classed as "yellow laterites," as a result of confusion. Age and Nature of Products. Locality Sulphuric Recent— red laterite {now forming) . . Kilanea Crater 1.57 per cent. Ancieut — red laterite Heeia 1.31 " Ancient— red laterite Waialua 1.68 " Ancient -red laterite Koliala 1.08 " Ancient red laterite Kokoloea 0.65 " Ancient -red laterite Paia 0.28 " Ancient — red laterite Makawao 0.04 " Mean of all Ancient red laterites 0.61 '• These red laterites vary in physical state from plastic to solid concretionary deposits. Four of these contain 80% of iron oxide, the alumina and silica having sejiarat- ed out. Age and Nature of Products. Locality. SuliJh'iric Recent — yeUow earth (now forming).Kilauea Crater 1.33 per cent. Ancient— yellow earth Hilo 1.29 " Ancient— yellow earth Laupahoehoe 57 " Ancient— yellow earth . . Ookala 0.71 " An' lent - Clay or Kaolin Koloa 5.85 " Ancient— Clay Wainaku 0.44 " Recent—Gypsum (now forming) ..Kilanea Crater.. ..44.73 " Ancient— ^'•ypsnm Koloa 55.81 " The ancient red laterites were all found to contain sul- phuric acid and certain of them, which have been pro^ tected from the leaching action of rain, contain even more than is found in the fresh laterites now forming und(»r th{^ action of the sulphurous steam at the Kilanea crater. We had persuaded ourselves that the layers of red laterite, lying at a depth of from two to four feet below the land surface, must have been deposited by the action of carbonic acid, derived from decaying vegetable mat- ter, on the iron in the soils. In most of the laterites, how- ever, no carbonic acid was found, jnid where the acid 53 was met with, it was in products where coral limestone was lying in the neighborhood. But the absence of car- bonic acid in the deposits or layers of laterite is no abso^ lute proof that the deposition was not caused by that agent. The carbonate of iron, when deposited, would part with the acid, and the iron would revert to the oxide form in which it is now found; the laterites thus distinctly differing in their behaviour from the concre^ tions known as "hardpans," one example of these con- cretions containing 30.26% Si O^; 5.83% Fe^ O3; 14.83% Alo O3; 13.61% Ca O; 8.84% CO.. In connection with the formation of laterites that are lying within two feet of the surface, we have to bear in mind that the red late^ rites formed, and forming near the active crater are little more than a foot below the surface, but are marked by crystals of sulphur, which attest their mode of formation. Nevertheless, we are still impressed that carbonic acid has also acted as an agent in causing the deposition of the laterites in given localities, although it has not yet been possible to obtain actual evidence to confirm the persuasion. Concerning the action of rainfall and temperature in the matter of laterite formation, the climatic conditions of the present time furnish no conclusive indications. The red laterites, and other disintegration products of the lavas, are found on the windward and leeward sides of the islands, and in the dry districts of Kau, Kohala, Paia, Waimea, and Honolulu, as well as in the wet dis- trict of Hilo, and the moderately moist districts of Ki- lauea, Koloa, Hana and South Hamakua. In concluding our observations upon the operation of chemical action in the disintegration of lavas on these islands, we refer once more to the extensive areas which 54 during earlier periods, have been the centers of vast vol- canic action. Lateral cones have been in action upon the slopes of the great craters, and over a large part of the surfaces of the islands. From these cones, tufa lavas were poured forth, saturated with steam and acid vapors, which caused the lava masses to undergo extreme altera^ tion in chemical composition at the time of ejection. Through fractures and fissures of the ejected lavas, steam, frequently charged with sulphurous acid, continued to escape, and these vapors carried on the alteration primarily caused in the tufas, and appear also to have operated upon areas of solid lavas, which Avere dis- charged without alteration, causing results in disintegra- tion that are not comparable with the effects of "simple weathering." The examination of the disintegration-products has not only furnished indications of the broad scale of areas over which chemical causes have acted in the decomposing of the lavas, they have led to the observation that the forma- tion of laterite has been, and is still being, due to the operation of the same chemical action. For whatever other causes or agents may have operated, it appears, without doubt, that the laterites of the Hawaiian Islands owe their origin, on a grand scale, to the action of sul- phurous steam in the disintegration of the lavas. If we extend our observations, and associate them with phenomena that have been noted in other countries, it is indicated that our conclusions may be found to allow of a more universal application. Authorities have been used to obtain a view of the prevalence of past volcanic action over surfaces of the globe, and we know of the basalt and tufa eruptions of Europe and the British Isles; of the lavas of India and Africa; and by the recent work 55 cf Eussel,- more of the magnificent areas of an earlier volcanic activity in the regions of Mexico and the Pacific Slopes. Further, we have noted the association of forma- tions of "laterite" with locations of igneous rocks; Lyell having shown the relations of laterite and basalts in Eng- land and Scotland; Blandford and Delesse the same con- ditions in India and other countries, and the "red ochres" of North-western America may be shown to have a rela^ tion to the volcanic movements of those regions. There^ fore the wide-spread appearance of these phenomena cause us to think that chemical causes may have ex- ercised an enormous and wide-spread action in rock- disintegration, and whose results may be still found re- corded in the soils of the older continents and lands. Weathering. — In previous paragraphs it has been im dicated that large masses, and areas of lava surfaces, have suffered primary disintegration under the action of chemical causes; and that other vast areas have not been acted upon by those special chemical agents. All lavas, however, whether they have, or have not been initially acted upon by chemical agencies, are resolved by the in- fluences of "simple weathering" into the palpable state in which they are called soils. The several influences that are understood under the term "weathering" may be summarily expressed as variations in atmospheric lieat and moisture. Great ex- tremes of temperature act powerfully in the fracturing and disintegrating of rock masses where moisture is present. In tropical climates the extremes of tempera- ture do not obtain, but the continuous action of high temperature, combined with rainfall, upon rock surfaces leads to the same results; and when these combined in- 56 fluences are aided bv the growth and decay of a hixuriant vegetation, the final work of" w, or yellow-brownish color, due to the hydration of the iron oxide. This effect compares, in a measure, with the results of oxydation under the action of steam upon the lavas, which, in the case of tufas, has been shown to produce colors of every hue and degree of vividness. These various results of weathering, under the action of different conditions of heat and moisture, will be further considered in connection with soils. At the time that the wi'itei- collected specimens of the great masses of solid lavas upon the several islands, specimens were taken of these same lavas that had been exposed to atmospheric action, and which represented states of disintegration under simple weathering. The condition of these weathering lavas is set forth by the following partial analyses. 57 WEATHERING LAVAS. Com- 1 Moisture. bined SiO.. FeO FeoO,, 1 AI„0., CaO water. Per cent. Per cent. Per cent Percent Per cent. Per cent Per cent 2.57 3.53 43 19 5 40 9 90 23.47 3.26 12.71 39.31 2 63 12.63 25 08 5.71 7.61 36 38 3 75 15 81 20.58 S 3 94 3.23 35.38 4.94 8.53 16.58 ^ 9.48 8 31 30.24 8 45 8.38 20.22 -s< 9.25 10.40 26.19 8.38 14.38 25.18 Mean. 5 70 7.63 35 11 5 59 11.27 21.85 7.06 The ayerage composition of these weathering lavas \v(; now bring into comparison, especially, with the hydrous solid lavas, with which they were originally identical, and with the tufas. The first line in the table is given to the non-hydrous solid lavas. Lavas. Mois- ture Per cent. 41 0.82 5 70 Com- bined water t SioJ FeO Pe.O;, AUO, CaO Solid (non-hydrous) Solid (hvdrons) Per cent. 13 1.96 7.63 11.49 Per Per cent. 1 cent 49.I2I 8.43 44 66 8.66 Per cent 4.04 5.62 Per cent. 21.94 17 06 21.85 23.58 Per cent, 9.24 8.23 35.11 .5.59 11.27 7.06 Tufas 7.72 26.60 1.69 25.79 1.41 In the individual analyses, as well as in the averages, the signal differences between the tufas and the normally weathered lavas, are seen in the exhaustion of the lime, reduction of the alumina, and enormous increase of the iron in the former, as compared with the steady loss of lime and increase of iron and alumina in the latter. The action of sulphurous steam upon the tufas has already been explained as a cause of the rapid removal of the lime, also of the alumina, from the lavas. The removal 58 of the silica, as the hydration of the lavas proceeds, will be spoken of in anotlier place, and in connection with the subject of the loss of materials from the land, as indicated by the examination of waters of discharge. In studying any given kind of rocks or lavas their characteristic features are brought out with a more im^ pressive clearness by bringing them into comparison with other classes of rocks. For this reason we shall give the chemical composition of various rocks which make up formations in North America, and from which the soils of those regions have been derived. For the composition of American rocks we are indebted to the incomparable collection of "Analyses of Eocks" by Messrs. F. W. Clarke and W. F. Hillebrand, of the U. S. Geological Survey. COMPOSITION OF AMERICAN ROCKS. Rocks. SiOa AI2O.. Fe.,0< Per Per Per cent I cent cent. (1) Siliceous Sandstones... 88.48 5.85 3 10 (2) Granites 72.501440 2 14 (3) Loess 67.761189 4.15 (4) Clay stones 57.00 24 50 1 .05 (5) Slates and Shales.. .. 51 36 15.54 4.14 (6) Iron Carb-silicates 36 00 122 2 52 (7) Lime Carb-sulf-silicates 4.42i 0.32] 0.30 FeO CaO MgO Per Per Per cent. cent. I cent. 0.00 00; 00 0.00 3.85! 26.45 44 1.76 3.65 3 18 1.82 4 44 0.00 50.65 Mean of Above 53 9312 00 6.81 American Basalts 49 . 15 15 66 9 . 52 Hawaiian Basalts 47 . 90 18 23 13 . 36 9 42 8.29 8.99 0.66 52 1.92 1 65 3 20 4.43 2.28 KoO Per cent 1 41 4.58 2 37 2 '81 Na,0 Per cent. 1.29 3.33 1.52 1 73 40 0.30 2 10 2.31 1.67 7.90 2.84 1 90 6 05 2 20 1.50 Our immediate purpose in com])aring the several classes of American rocks with Hawaiian lavas will be very presently seen in the discussion of Hawaiian soils. There is a profound geological reason, however, for the comparison of tlicse classes of rocks with basalts. If we 59 refer back to the decomposition-products, which we have shown are severally formed from the disintegration of layas, we are struck by the resemblance of these Hawai- ian products, in chemical composition, to the respective kinds of American rocks. At this time we do no more than note this analogy, and shall reserve its discussion until the time when the whole subject matter of this study shall be considered at a later time in a work of greater scope and detail. HAWAIIAN SOILS. Under tlie action of the several agents and modes of disintegration, it has been shown how that the lavaa have become decomposed and resolved into earths. It is thus from the decomposition of the lavas that the soils of these islands have been derived. It has been shoAvn moreover, that in the course of dis~ integration the lavas have yielded a series of earths and products, each of which has been more or less separated from the others, and many of which have been removed from the place of their formation, and have gone to the laying down of other formations, either at lower levels on the land, or under the sea. Due to the small areas of these islands, and to the acute declinations from the mountains to the ocean, and also to heavy rains, especial- ly on the windward exposures, the separation-products in the disintegration have been borne largely to the sea. If the lavas, in the course of disintegration, have fallen into these several classes of decomposition products, many of which have been separated from the mass, and carried away, it then appears that the soils in i)lace must bear only a remote relation to the rocks from which they have been derived. A comparison of the composition of the lavas, with the average of some six hundred soils de- rived from tliem, indicates that the relation is a very dis- turbed one. In the following comparison the constituents 61 of the soil are calculated on the mineral matter, iu order to note the extent of the alteration. COMPOSITION OF HAWAIIAN LAVAS AND SOILS. Material. Si FeO + Fe2 O., AI2O3 Ca MgO Na,0 K, Hawaiian Lavas Hawaiian Soils. Per cent. 47.90 27.54 Per cent. 13.36 36.45 Per cent 18.23 22.64 Per cent. 8.99 0.46 Per cent. 6.05 1.07 Per cent . 2.20 1.19 Per cent. 1.50 0.62 In general, the comparison shows that the silica, lime, magnesia, the alkalies, also much alumina have been borne away. Fortunately the life in the sea has gathered up much of the escaping lime, and restored it to us, at our very doors, in the form of coral reefs which begirt the islands. The full significance of the vast difference in the composition of the lavas and soils will be con- sidered in detail at a later time, the X3resent purpose being only to illustrate that these, and all other soils bear only a distant relation to the rocks from which they have been formed. There are formations and soils which represent the separation products, and other soils the residual products of the lavas and rocks from whose disintegration they have been derived. We have to deal, in the main, with soils that are the residual products of disintegrated lavas. Attention has alreacty been called to the action of tem^ perature and rainfall in the matter of weathering of rocks, and in the final disintegration of the materials forming soils. The variations in rainfall and the effects of less or greater precipitation, have been such as to ad- vise the division of the soils into uplands and lowlands, the former representing the areas of larger, and the latter 62 the areas of smaller rainfall. lu later para<>raplis we shall discuss these evidences of the direct results of local climatic conditions. Just now, we have to consider our soils from a point of view from which it may be seen that the great differences in their nature, and economic value, are due to other causes, no less than to variations in climatic conditions. The soils of these islands, and of other countries where the lands bear a resemblance to these, pass generally under the definition of red, or yellow soils. At first view this definition seems to be in place; but when we come to examine into the origin of the different reds and yel- lows, and note their chemical compositions, it is then found that this general definition does not apply, and may actually conceal the most basic and vital differences. Moreover, the definition is too definite, giving the impes- sion, more or less, that all our land surfaces are either vivid reds or bright yellows, which is far from the case. The red soils vary in colors from dark reds to crimson and light reds, the latter usually being contiguous to, or found within areas where soils that vary in color from light to reddish-yellow prevail. The dark, blood-red soils are frequently more distinctly marked off from those of other colors; but this is not, by any means, al- ways so. Then, in addition to red and yellow soils we shall have to speak of certain dark soils, and of their origin and characteristics. The common definition has been carried still further, and used to denote the qualities of the soils : In general, the red soils have come to be considered as good soils, and the yellow as poor soils. This also is far from being wholly correct; although certain red soils are very fertile, and many yellow soils are sterile. Actual experience has 63 shown men engaged in the island agriculture that whilst given areas of red land are rich, and permanent in fertili- ty, there are other red soils in which, practically speak^ ing, nothing will grow. This is actually the case, despite the general definition which has marked the red soils as good soils. We have therefore, in the first place, to inquire into the causes of difference between the kinds of red soils, and into the enormous differences in their fertility and economic value? Dark Red Soils. — In the previous paragraphs, which were occupied by the examination of the several forms of lavas, the results and conclusions arrived at were sum- med up as follows : "Amongst Hawaiian lavas are those which have discharged from craters, flowing and cooling into rocks having the composition of normal basalts. Others, originally of the same composition, have under- gone such alteration that they now compose rocks masses having a radically different chemical composition and color appearance. This alteration took place at the time of ejection, and under the action of chemical causes, and previous to the later action of secondary causes of rock disintegration, such as 'weathering,' which has apparently been the only agent of disintegration of cer- tain of the normal lavas." There are broad and defined areas, especially upon the Islands of Mauai, Oahu and Kauai, where the lava dis- charges from the great craters have flowed and cooled into rocks, and upon which simple weathering appears to have acted as the only influence in their disintegra^ tion; and there are other more acutely marked areas where the lavas have undergone such alteration at the time of ejection that their appearance and composition are radicallv different from those of normal lavas. (U The areas Avliicli represent tlie flows of normal, unal- tered lavas, whose surfaces appear to have become disin- tegrated by the slow action of simple weathering, are found, in the greatest part, upon the lee sides, or south exposures of the Islands; and consequently within the districts of smallest rainfall. This is far from being exclusively, although it is mainly so. It happens fur- ther, that whilst certain of these areas are not exposed strictly to the south, the rainfall, due to local topo- graphy, is nevertheless small. In distinction from what we have said is chiefly the rule, there are at least two small and indistinctly defined areas on Hawaii where it appears that the underlying lava has come down from the great craters by a free flow, and has been largely free from other causes of disintegration than weathering. Those areas have a full windward exposure, and lie under, the one a very heavy, and the other a medium rainfall. The areas, that have been defined as representing dis^ charges of more or less free flowing, normal lavas, which lavas have undergone slow disintegration in hot expos- ures, with a minimum rainfall, are now the districts marked by the ])r(Mlominance of tively derived. In these comparisons are given only the more dominant constituents — silica, iron, alumina, and lime, and these are presented on the basis of mineral matter, free from moisture and combustible matter. COMPARISON OF LAVAS AND SOILS. Materials. SiO, Per cent. A.— Normal Lavas fSolid^ 47.59 Normal Lavas (Weathering) [40.35 Dark Red Soils 31.45 B. -Tufa Lavas Yellow and Light Red Soils 32.84 23.63 Fe,0.. Per cent. 15.02 20.52 29.78 33.92 43.11 AloOs CaO Per 1 cent. 19.92 25.23 26.94 29.11 18.33 Per cent. 8.88 8.11 0.57 1.74 0.37 82 In the passing over of the solid normal lavas, throiii;h the state of "weathering/' into the dark red soils, is noted a most p'adual ehani>e in tlie relations of the elements to each other. In the weathered specimens the relation of iron and alnmina to each other is the same aft in the solid lavas; and although the alnmina has been, in part, removed and the iron has accumulated, in the course of the final resolution of the lavas into soils, the relation of the two chief constituent elements has been, visibly maintained. In the tufa lavas it is seen that a most violent dis- turbance of the relations of the elements was effected, and largely before the action of weathering began. As these lavas passed finally into soils, the relation between the iron and alumina underwent a change that complete- ly reversed the quantitative proportions of those elements as they existed in the original lavas. This change has also removed the ground of comparison, and rendered the yellow and light red soils extremely dissimilar in texture, color, and composition to the dark red soils derived from the normal lavas. We do not lay great stress upon the differences in lime and other soluble elements between the two types of soils, since climatic conditions, to which we shall presently refer, have had a great determining bearing upon the pro])()rtions of those elements left in the soils. It has been found, however, that in instances where yellow and dark red soils have been formed near together, and the rainfall essentially the same, the dark red contain, uniformly, more than doubh' the amount of lime found in the yellow soils, and it was also found that in the yellow soils the largest proportion of the lime was contained in siliceous combinations in the "in- 83 soluble residue," and only recorded by the "absolute analysis." We have, in previous paragraphs, dwelt upon the sul- phuric acid as an indicator of the mode of disintegration, and of the formation of soils. Like lime and the alkalies, this acid is readily removed by rain when in combina^ tion with the bases named, and is thus found in least quantity where the rainfall is the greatest. The upland soils contain 0.166%, and the lowlands 0.210%, of sul- phuric acid. The "dark red soils," which have been formed in dry conditions, contain 0.332%, and the "yellow and light red soils" 0.140% of sulphuric acid. If we leave the general averages, and note the variations in sulphuric acid in some soils of the same locality, where the rainfall is the same, a new set of indications present themselves. In the following data is found the measure of variation in the soils of localities, where the variation cannot be due to superficial causes such as rainfall. Localities, Heeia, Island of Oahu Honohihi. Island of Oahu Hilo, Island of Hawaii Ookala, Island of Hawaii Mean of 820 Soils of America and England Minimum S O, in Soil Per cent. 0.050 0.050 0.210 0.120 Maximum S O3 in Soil Per cent. 0.828 0.650 1 290 0.450 0.04 Per cent. In the cases where a high content of sulphuric acid has been found, it appears largely in combination with fer- rous iron, causing conditions inimical to plant growth, which we shall speak of again. These high amounts of sulphuric acid have, so far, only been found in soils where great volcanic activity, and special vapor action on the 84 lavas, liave transpired. The small amounts of the acid found in soils within the same locality, and not half a mile apart, is to be accounted for, we believe, by the circumstance that we have already explained, viz: — "there are areas, contiguous to such as have been acted upon b^^ sulphurous steam, where no traces of steam action are evident." "Poisoned Spots."— In addition to the typical differ- ences in the soils, which have been traced to the different lavas from which they have been derived, there are yet other variations. Within broad areas where the soils have been derived from solid rocks by weathering, places or patches are found which appear, and behave quite differently from the surrounding land. Hardly anything will grow on some of these patches, and the^^ are called "poisoned spots." In some districts upon an area of 1500 acres, patches summing up to eighty or one hundred acres may be found. We are persuaded that these poisoned spots were caused by the action of sulphurous steam upon the solid lavas after their emission from the crater. We are led to this view by what is observed to- day at Kilauea over the area of the crater floor, and which we have described in earlier paragraphs. At the active crater we noticed the action of the acid steam in changing the appearance and composition of the lava. We also noticed an accumulation of sulphuric acid in the altered lava, it having increased to three times the amount found in the lava where no steam action was proceeding. In comparison, these "poisoned spots" in the fields present a color appearance Avhich not only causes them to resemble the steamed patches on the crater floor, but distinguishes them from the land around. The poisoned spots are most numerous in districts on 85 Hawaii, biit tliej are also found on all the islands. On Kauai, in the district of Makaweli, the superb breadths of "deep red soil" are spotted with these patches, which are yelloAV to light red in color, and from one-half of an acre to more in dimension. Near to the old crater, which is now used as a large reservoir, these yellow colored spots are very acutely defined. Samples of the soil were taken from three of these poisoned spots, two from Hawaii, and one from Oahu, in each one of which the sulphuric acid was above 1.0^/c, and the average of the acid in the three soils was 1.33%. These data leave us almost without doubt that in the origin of those "poisoned spots" sulphurous steam was one of the potent factors. We now present an example of a soil, of a gray-yellow color, in which the cause of color is not so evident as is the case with the yellow and light red soils generally. At Makaweli there is a breadth of land known as the "yellow ridge" Avhich runs through a broad area of dark red soil, but is most acutely defined from the latter. This "yellow ridge" is a narrow strip running down through the main area of dark red soils which distinguish the locality. The lava which formed the yellow ridge was a later and distinct flow, but it came from the crater from whence the lavas forming the dark red soils came. Samples of the lava, in a state of weathering, which i» forming the dark red soil, and also of the lava forming the yellow soil, were taken. The one lava, on its weather- ed surfaces, is already deep red; the other lava a dull yellow. The analyses of these two lavas gave as follows: 86 \ Lavas- Com billed water Per cent 4.85 7.58 SiOa FeO Fe^Oa ALO;, CaO Red Colored Yellow Colored Per cen'. 32.00 3332 Per cent. 380 2.81 Per cent- 14.78 16.04 Per cent. 13.76 13.23 Per cent. 8.73 9.00 We have said that these hivas "were from the same crater," and the analyses say that, in clieniical compo^ sition, they are intrinsically identical. Yet the one is red and the other yellow! We note that the yellow con- tains nearly double the amount of combined water, and its iron is more advanced in oxydation, than is the case with the red lava, which is older than the yellow lava. These are the only marks of difference; but they are the peculiar marks of difference that would result from the action of steam at the time of, or after ejection. That steam acted upon the lavas after emission in this locality is shown by the "poisoned spots" in tlie red soils near by. But there is no excess of sulphuric acid in the yellow lava, or its soil, to indicate the past action of sulphurous steam. We remember, however, that at the Kilauea crater, at tliis time, neutral steam is escapin**' throunh the lavas, and tliat in places where the steam was acidic two years ago, it is neutral to-day. These memoranda cause us to think that the yellovv' ridge lava, and con- sequently the soil, owe their color to the action of neutral steam at the time of, or after ejection, and that the prac- tically unaltered state of the yellow lava may be due to the absence of sulphuric acid in the steam. The soils from these lavas are as follows: 87 Soils Com- bined water etc. Fe.O:, AI0O3 OaO Nitrogen Dark Eed Soil Per cent. 11.36 12.44 Per cent. 28.73 26.93 Per cent. 26.59 23.53 Per cent 0.338 0.160 Per cent. 0.171 Yellow Kidge Soil 0.165 The relations of the irou and alnniina indicate a ten^ deney in change resembling the differences between the red and yellow soils in general. The lime points dis- tinctly in that direction. The larger amount of water of combination in the yellow soil corresponds with what was found in the lava; and the nitrogen content indi- cates that the excess of combined water in the yellow soil is in the form of mineral hydrates, and does not come from any excess of humus or organic matter, an indication which is supported by the exact sameness of the rainfall under which the dark red and yellow soils, in this example, were formed. This example is given in order that it shall be understood that there are instances of yellow soils whose origin and causes of formation are not yet fully understood; and which, with our present knowledge, we cannot safely include within any type that has been established. UPLAND AND LOWLAND SOILS. So far, the soils have been considered from the stand- point of the lavas, and of the processes of disintegration, by which these have been resolved into the materials which form the mineral constituents of soils; and as a result of the great differences in the soils, due to dis- similar causes of disintegration, they have been con- sidered under several more or less definite types. 88 Soils, liowc'ver, are not merely decomposed rocks: They possess the elements which constitute the lavas, but t\u'\ contain soinethinii- more. We have seen that Hawaiian soils hold as high as 20 per cent, of "combus- tible matter," and while some soils have less, others contain more than this amount. This combustible mat- ter consists, in part, of simple water that has combined with the elements of the lavas, in the course of disin- tegration, and is driven off by heat, A large part, how^ ever, is vegetable matter, and organic substances that result from vegetable decay. These vegetable materials, as our studies of the lavas have shown, are not found in essential rocks, and the elements that form these mate- rials are seldom found in rocks, excepting such as are of organic origin. Carbon, nitrogen, hydrogen and oxygen are the elements which bear the burthen of all vegetable structure. Carbon and nitrogen were primarily derived from the air, as they are being obtained from that sphere to-day. Hydrogen and oxygen, in addition to taking- part as individual elements in the structure of plants, combined in the form of water they not only assist in conveying other elements from the air to the soil, but, as water, they are absolutely essential to the life and growth of plants, and out of all proportion to the actual amounts that take \)i\vt in ])lant structure. Scientists have already shoAvn how many tons of water have to be absorbed and evaporated in order to produce one ton of wheat, barley, oats, etc.; and we have just determined the weight of water used by the sugar cane, during a j)eriod of several months, in the making of one pound of its own substance, and that weight is enormous I These examples are not necessary however, to prove what is said: We need merely consider the great 89 deserts, oii the one hand, and on the other, the vast areas of vegetable luxuriance over the face of the earth, and it is at once apparent that vegetation and rainfall are concomitant conditions. As the combustible organic matter in soils results from vegetable decay, and the amount of vegetation ia proportional to the rainfall, it then appears that humid soils, or those formed under great rainfall, should con- tain more organic matter than soils formed in arid con> ditions, and under a minimum rainfall. Further, as nitrogen is a constant, and only slightly variable factor in the organic matter, the nitrogen content of soils in regions of great rainfall should be uniformly higher than in the soils of dry districts. On the other hand, and in acute distinction from the element nitrogen, which is brought to the soil as a direct and indirect result of rain- fall, the several mineral elements of the soil, such as silica, alumina, lime, etc., which were originally there, and which can be taken away by water, are liable to be the lowest where the rainfall is the greatest, and the highest where precipitation is least. In view of these considerations, the agricultural anal- yses of our soils during the past three years were resolv- ed into the two classes — "upland" and "lowland" soils. These analyses have covered something over 1,300 sub- samples, in the most of which only the lime, potash, phos- phoric acid, and nitrogen were determined. In all cases the samples were taken personally by the writer, or in fields, and by methods, stated by him. By "lowlands" is meant the land areas which run from the sea to an elevation of about 500 feet, and by "upland," the upper cultivated areas, which rise to about 1,500 feet. The appli- cation of these terms, however, is controlled by the dis- 90 trict: For example, iu parts of Hamakua the lowlands commence upon bluffs that stand 200 feet or more above the sea; whilst in llih* the rainfall is so ive a brief summary of our tindiniis during the past three years, which have already been discussed and publislicd in some detail. The data represent the "lowlands'' and "uplands" of Oahn, Mani, parts of Ha- waii, and Kauai, and distinguish more or less acutely between areas of small and greater rainfall. Soils. Lime. Potash. Phosphoric Acid. Nitrogen. Uplands Lowlands Per cent. 0.331 0.471 Per cent 0.297 0.328 Per cent. 238 0.213 Per cent. 0.465 195 These data represent the average of virgin and crop- ped soils on the "upland" and "lowland" areas. The lime contents in the respective soils support, without anv question, the original hypothesis "that the lime, etc., would be liable to be the lowest where the rainfall waa the greatest." In the matter of organic matter and nitrogen onr findings are still more emphatic in pro- nouncing that "vegetation, nitrogen and rainfall are concomitant conditions." We can go one step farther for illustration of this relation of the nitrogen content of the soil to the rainfall: Upon all the four Islands most enterprising efforts are being made to extend the coffee production. The lands suited to coffee growth are at the elevation where it ceases to be ])rofitable to grow sugar, or from 1000 to 2000 feet above the sea level. We have analyzed soils from some of these lands, and have 91 obtained analyses made by other chemists, covering the coffee districts on Hawaii, and prospective coffee lands on ]Maui. The nitrogen in these thirteen coffee soils we compare with the nitrogen contents of the cane lands. Soils Approximate Mean Elevation. Approximate Mean Rainfall. Nitrog-en in Soil. Lowland Caue Soils Upland Caue Soils Coffee Soils Feet 300 900 1,800 Inches 50 90 130 Per cent. 0.195 0.465 1.237 We are indebted to the Government records for the "approximate mean rainfalls," and to our own observa^ tions in given districts. In one of the districts on Maui, a series of rain gauges placed at intervals up the moun^ tain side, from 200 feet to 3000 feet, gave an annual rain- fall of 28 inches at 200 feet elevation; 60 inches at 900 feet; and 179 inches at 2800 feet. In the Nuuanu valley, Oahu, the rainfall at sea level is near 30 inches, and at 900 feet it is 118 inches. On account of given questions and difficulties bearing upon the maintenance of fertility of the upland soils, as compared with the lowland soils, a very exhaustive seriea of experiments was undertaken in order to try to obtain precise data concerning the "availability of the elements, of plant food in the upland and lowland soils," the results of which will be set forth in the second part of this pub- lication. The soils used in these experiments were, in the first place, subjected to exhaustive analysis in order to be able to note to what extent any differences in be- haviour of the upland and lowland soils, under the action of solvents used, would appear to be due to fundamental or structural differences in their composition. These 92 analyses we introduce at this place on account of their special bearing also upon the question of the relation of rainfall to the organic matter and nitrogen content of soils. The analyses include and represent the average of nine type samples and 108 sub-sam])les of upland soils, and of the same number of t3'pe and sub-samples of low- land soils. As already remarked, the samples were either- taken by the writer, or under his (liv(M't instructions. Districts of the four islands are end)raced, and the climatic conditions and length of time that the lands have been under cultivation, have been carefully recorded. The results of the analyses are as follows: LOWLAND SOILS. Constituent Elements. Insoluble Matter, Moisture Combustible Matter Si 0-2 (insoluble). . . . Si 0-2 soluble) Ti O2 P'^05 SO3 00-2 Fe2 O3 Al. O3 Ca O Mg O Mn3 04 K2 O Na-iO Agricultural Analysis. Per cent. 35.150 9.031 15.460 1.780 396 0.234 0.290 19 980 16 155 390 0.802 0.187 0.286 0.355 99.681 Insoluble Residue. Insoluble in HCl. Per cent. 27.440 35.000 8.060 0.770 8 290 14 980 1 090 1 250 0.890 2 510 100.280 Absolute Analysis. Water-free Soil. Per cent. 16.804 10.290 13 391 5 070 724 0.175 0.202 25.150 23 539 0.851 1 349 0.124 650 1 344 99.663 Mineral Matter. Per cent 12.569 16 049 6 190 870 210 30 220 28 292 1.022 1.621 148 0.781 1 615 99.587 Acidic Elements in the Soils - 35 888 Per cent Basic ■ '• " 63 689 Total Nitrogen in the Soils (' 291 Soluble in 3 per cent K O H, in 30 hours - 0.204 Insoluble " " " 087 93 UPLAND SOILS. Constitue t Elements. Agricultural Analysis. Insoluble Residue. Absolute Analysis. Insoluble Matter Per cent. 27.870 12.290 20.600 Insoluble in UCl. Water-free Soil. Mineral Matter. Combustible Matter Per cent. Per cent. 23 300 10 674 9.903 5 192 869 0.128 030 26 174 20.059 640 0.945 0.153 0.710 1 280 Per cent. SiO'2 (insoluble) 33.450 31 230 9.630 1.050 14.068 Si 0-2 (soluble^ . . 13 042 Ti02 1 840 0.470 157 030 21 810 13 621 294 610 187 0.272 391 6.739 P.2 O5 1,132 S 3 0.166 C O2 "3 740" 14 640 0.950 790 "1.300"' 2 630 Fe2 O3 34.120 AI2 O3 26 152 Ca Mg 0.834 1 193 Mns O4 K-i 0.200 0.925 Na2 1.668 99 953 99.410 100.057 100.087 Acidic Elements in tbe Soils = 35 . 147 Per cent. Basic " " " =64 940 Total Nitrogen in the Soils = 490 0.347 0.143 Soluble in 3 per cent. K O H, in 30 hours Insoluble " " It is uuderstood that these analyses are of soils derived from the same lavas and lava flows; so that any dif- ferences indicate the measure of the action of climatic conditions, the upland soils showing the effects of greater rainfall. Ver}^ noteworthy is the difference in the relative pro- portions of soluble and insoluble silica. In the upland soils the insoluble is the greater, whilst in the lowland the soluble silica has accumulated. The enormous amount of soluble silica in all the soils will be spoken of again. A notably greater increase of iron is seen in the upland soils, with a reduction of the alumina. In the lowlands, the lime is appreciably higher than in the 94 uplands, whilst llic iihosplioiic ncid is just liic reverse. But from the analyses it would not appear that there is a difference that could seriously affect the fertility, since the upland soils contain an ample total amount of the elements of fertility for an indefinite length of time. The question, however, which concerns the immediate crop is not the total, but the (iraihihlr amount of the elements indispensable to fertility. In the second part of this work the relative state of availability of the elements in the upland and lowland soils will be fully considered. Finally, it is noticed that, in respect of the combustible organic matter and the nitrogen contents of the upland and loAvland soils, these tables of analyses fully support the data already furnished in showing that these soil constituents are in proportion relative to the rainfall. The nature of the nitrogen compounds in the soil will not be considered in detail at this time. So far, the data show that 70.4% of the total nitrogen in the soil is solu- ble in a 3% cold solution of potassic hydrate, under an action of 30 hours duration. Concerning the iratcr- soliihlc nitrogen, it is indicated that this is largely in the form of amido-acids. Qualitative tests, in which the soil was digested for 30 hours at a temperature of 4.") degrees centigrade, and the solution treated with mercuric nit- rate, gave ample indications of amido bodies being present. This mattei- is of signal importance on account of its bearing u]»on the question as to the form (»f the nitrog(Mi in which i>lants, growing in acid soils, take up their nitrogen, a (juestion to which we have already re- ferred, and to which we shall recur. In presenting our observations ujxtn the relation of rainfall to the oi-ganic matter and nitrogen in soils, we have to call attention to the fact that other observers 95 have reaehed conclusions of a directly opposite nature. Professor Hilgard ("Kelatious of Soil to Climate") fur« nislies data showing that the "humid soils" in the United States contain less organic matter than the "arid soils." His conclusions result from the averaging of not less than 779 soils, and, on account of the reputation of the author, require a very careful attention. In recapitulating the conclusions set forth in the pre- vious paragraphs we have, with the aid of our present knowledge, been led to consider the types of Hawaiian soils as follows: A. — Geoloijivnl ( la-ssificafion. 1. Dark Eed Soils. — Soils formed by the simple weathering of normal lavas, in climatic conditions of great heat and dryness. 2. Yellow and Light Red Soils. — Soils derived from lavas that underwent great alteration, under the action of steam and sulphurous vapors, at the time of, or after emission from the craters. 3. Sediimentary Soils. — Soils derived from the decom- position of lavas at higher altitudes, and the removal and deposition by rainfall at lower levels. It has already been said that there are soils which can not with certainty be placed under any one of these three types. B. — Climatic Classification. 1. Upland Soils. — Soils formed under lower tempera- ture and greater rainfall, and distinguished by a large content of organic matter and nitrogen, and by a lo^v content of the elements of plant food in an available state; these elements having been removed by rainfall. 96 2. Lowland Soils. — Soils formed under biglier temper- ature and smaller rainfall, and distinguished by a lower content of organic matter and nitrogen, and by a higher content of the elements of plant food in a state of imme- diate availability, which is due in part to the receipt of soluble constituents from the upper lands, and to a smaller rainfall over the lower levels, RELATIVE FERTILITY OF THE SEVERAL TYPES OF SOILS. The "dark red soils," found in districts already de- scribed, and the "sedimentary soils" are more or less, and almost uniformly, fertile soils. The "yellow and light red soils" are not marked b}-^ anything like the same uniformity in character and fertility. Certain of these, when first brought under cultivation, produce good crops, but the source of fertility is not permanent. When tw^o or more crops have been removed the power to produce weakens greatly, and the restoration and maintenance of fertility is difficult, special treatment being required. Others of these soils, marking small or larger areas near the centers of a past great volcanic activity and chemical action on the lavas, from the first are not productive. In three special loca- tions the managers of ])lautatious have respectively said to the writer "the cane will not grow in tliis bright red soil." "Wherever that light red soil is mixed in with the other the land is poisoned." Further, "when the cane roots strike that (dd brown-yellow stuff they turn up and will not face it." Now, in each of the soils from the three locations quoted we have found more than 1.0 per cent, of sulphuric acid, Avhich is in combination with 97 tlie low oxide of iron, tlins producing a compound actual- ly poisonous to plant life. We had suspected the presence of these poisonous, iron comj)ounds, and the special examinations have con- firmed the supposition. Knowing, however, what certain of the fundamental reasons of non-productiveness are, we shall now be able to grapple with the trouble, and, in time, neutralize the causes of sterility. In order to furnish a more definite idea of the relative productiveness of the different types of soils, we shall present certain data now in our hands. We shall first explain that a system of control is in use by which the production of each sugar plantation upon the four islauda is known. Each plantation furnishes to the Bureau of this experiment station an annual statement, showing the number of acres of cane manufactured, and the sugar produced. The writer, by reason of the repeated visits to each plantation, and his knowledge of the type or types of soil on each plantation, has been able to arrange the soils of each plantation under their proper types, and to attach to each its actual production. In the following table are included data only from plantations where soils of a definite type or types are found, and where the conditions of water-supply and temperature are com^ paratively uniform. Great variations in these physical conditions could upset any comparison based upon variations in the types of soils. The data are not a state- nient of one crop, but represent the average of the past three years, and are therefore free from any incidental deviations from the mean. The data bearing upon the three types of soil are as follows: 98 Types of Soils. Dark Red Soils Yellow aud Light Red Soils ■Sedimentary Soils Approximate ^o. of Acres. 30,000 32,000 20,000 Yield of Sugar Per Acre. 10,411 lbs. 6,291 " 10,301 " The cane crop on these ishiuds, aUowing for the time of prepariuj;' the i'round before phmtiug, is two years iu production, and two crops are always in course of grow^tli at tlie same time. The "approximate number of acres" given allows for this consideration, w^hich implies double the area required by the crop taken off in any one year. The land areas not embraced iu this table would fall chiefly under the type of the "yellow and light red soils" Avere it possible to include them. It is seen that the "dark red" and "sedimentary" soils are vastly more productive than the "yellow and light red" soils. One feature in favor of the latter is the circumstance that whilst they produce much less in quantity, the quality, taking the cane crop ag an example, is distinctly higher. This is exemplified by the following table, wiiich gives the production of certain sedimentary and yellow soils from several districts. Soils. Tons of Cane Per Acre. Purity of the Cane Juice. Tons of Su- gar per Acre. Tons of Cane to One Ton of Sugar. Sedimentary Soils 47.80 23.60 Per cent. 84.2 90 5 5.11 3.08 9 1 Yellow or light Red Soils. . 8.1 In each example used in this comparison the sedimen- tary and yellow soils are upon the same plantation; con- sequently the manufacturing data are strictly compara- tive, the same mill and mode of treatment being used 99 upon the- cane from the two kinds of soils. Neverthe- less, it is apparent that the different types of soil vary greatly in economic productiveness. The yellow and li*»ht red soils, by special treatment, can doubtless be increased in fertilit}^ They cannot, however, in our opinion, be made equally productive with the other types. But for the good management, and very conspicuouf»» economy practiced upon certain plantations where the yellow and light red soils predominate their position would be other than it is to-day. Comparison of Hawaiian and American Soils. — At the close of the examination of the Hawaiian lavas, we brought these into comparison with the kinds of rocks found in North America. It will now be of interest and value to compare the soils derived from the American rocks with the soils of these islands, which we have dis- cussed. A great geological interest will be found in this comparison, and its importance will consist in enabl- ing us to grasp the significance of striking variations, and in guarding us against applying conclusions drawn from the study of given soils, formed under their own characteristic conditions, to other absolutely different types of soils, which have resulted from rocks and con- ditions of quite dissimilar kinds. For the purpose of the comparison, we have examined the soil analyses published by the several experiment stations of the United States, and have received copies of analyses made by the laboratory of the United States Department of Agriculture, Washington, under the direction of Dr. H. W. Wiley, through whose courtesy they have been placed at our disposition. Accompanying the analyses. Dr. Wiley writes "unfortunately a large part of the soil analyses available have been made accord- 100 ing to different methods, and are not comparable. I send you the analyses of twenty different soils which were made in this laboratory by the same method." These analyses furnished by Dr. Wiley were the usual "agricul* tural analyses," and therefore did not furnish data for a geological comparison with the "absolute anal^^ses" made of Hawaiian soils. In the absence of data corresponding to our purpose we sent to thirty experiment stations in the United States asking for small samples of soil from the experi- ment fields to be sent to us. Most of the stations respond- ed to our request, including the States of Rhode Island, New Hampshire, Maine, New Jersey, Michigan, Illinois, Indiana, Ohio, Wisconsin, Kentucky, North Carolina, Missouri, Tennessee, Texas, Nebraska, Iowa, Minnesota, North Dakota, Idaho, Washington. Samples from otheT States were received, but too late for use in the examina^ tion. It is thus seen that the greater part of the soils received are, geologically, very largely representative of the American regions of glacial action and "drift" form- ation; and further, that they are mainly derived from rocks, whose origin was anterior to the laying down of the carboniferous formations, admixed with others, ovei> given areas, which had a volcanic origin. It is in nowise claimed that these represent fully American soils. As only an average composition of these American soils was required, one sample Avas made up by taking an equal weight of each sample received from the twenty States; the samples, instead of the results of separate analyses, being averaged. The analyses were made pre- cisely the same as those of the Hawaiian soils. The re« suits are as follows: 101 ■ COMPOSITION OF AMERICAN SOILS. Constituent Elements. Insoluble Matter Moi-ture Combustible Matter . . Si O2 (insoluble) . . . . Si O2 (soluble) P'iO-S S O3 C 0-2 Fe. AI2 Oe Ca O Mg O Mns O4 K-2 O Na..2 Agricultural Analysis. Per cent. 84.890 1.950 5.077 0.192 0.072 0.035 6.230 0.255 0.354 0.260 0.237 0.213 99.762 Insoluble Residue. Insoluble in H CI. Per cent. 79.540 8.580 0.200 7.200 0.610 0.440 1.500 2.200 100.270 Absolute Analysis . Water-free Soil. Per cent. 5.180 69.017 7.393 0.343 0.073 0.037 12 561 0.787 0.731 0.265 1.530 2.114 100.031 Mineral Matter Per cent. '72.697" 7.840 0.361 0.077 13 250 0.830 0.771 0.278 1.622 2.229 99 954 Acidic Elements in the Soil = 81.014 Per cent. Basic " " = 18.9S0 Total Nitrogen in the Soil = 0.219 Absorptive power of the Soil = 48.4 " The statement of analysis puts before us a type of soils fundamentally different in structural composition from Hawaiian soils. In the first place, is noted the relative state of solu- bility of the constituents in the two soils, which is set forth by the "agricultural analysis." The relative pro- portions of the respective soils that were found soluble by warm digestion with concentrated hydrochloric acid were as follows: Soils Soluble. Insoluble. Hawaiian Soils American Soils Per cent. 68.894 15.110 Per cent. 31.106 84.890 102 This enorinoiis differeuce in tlio relative proportions of the two soils that yield to the solvent action of strong hydrochloric acid is due to the fundamental difference in structural composition. The American soils are high- ly acidic, and the Hawaiian ultra-basic in constitution, which is shown by the following comparison: Soils. Basic Constituents. Hawaiian Soils. American Soils. Per cent. 63.717 18.980 j Acidic Constituents. Per cent. 36.458 81.014 A more extreme difference in the fundamental strvic- ture of soils than is set forth by these data is not con- ceivable. The comparatively small action of strong hydrochloric acid upon the American soils is amply ex- plained. The proportion of bases in those soils is 18.98%, or less than one-fifth of the total mineral matter; whilst the proportion of the mineral matter soluble in the strong acid was 16.25%. The nature of the great variation in the proportions of bases in the soils of the United States and of these Islands is set forth by the following com- parison : Soils. Fe, Ah Oe CaO Mg KjO Na^ Hawaiian Soils American Soils Per cent. 59.240 13.250 Per cent. 698 0.830 Per cent. 1.242 0.771 Per cent. 737 1.622 Per cent. 1.420 2.229 These data require very little immediate comment. It is shown that the cardinal difference lies in the relative contents of iron and alumina. Concerning the lime and 103 magnesia it is less important to note the similarity in amounts than to remember the notable difference in the state of solubility in the respective soils. The compara- tive amounts of potash and soda in the American soils indicate matters of geological interest, as well as of agricultural moment, to which we shall recur. The small proportion of basic elements in the Amer- ican, as distinguished from Hawaiian soils, involves a high content of acidic constituents, the bulk of which is composed of silicic acid, or silica. In drawing a com- parison between American and Hawaiian soils we note not only the relative difference in total amounts, but also in the proportions of "soluble" and "insoluble" silica, and this latter difference has a very profound agricultural and geological significance. The following figures give the relative amounts of soluble and insolu^ ble silica. Soils. Soluble Silica. Insoluble Silica. Hawaiian Soils American Soils Per cent. 15.308 7.840 Per cent. 12.619 72.697 The mineral matter of the American soils contains no less than 80.537% of silica, only 7.840% of which is asso- ciated with bases that are soluble in strong hydrochloric acid. The total silica found in the average of more than 1300 sub-samples of Hawaiian soils is merely 27.927%. And of these 27.927 parts, only 12.619, or actually U% of the total silica, are insoluble; the greater part being combined with the bases, and is set free by the action of the strong hydrochloric acid. Tlie soluble silica in the 104 H:nv;iiian soils is just (loiiblc tlic aiuoiiut loiiiid in tlu- AiiHMican saiu])U's. AjiTiculturally this is in-obably of hiji'li iin]toiianco, especially in the matter of the cane crop, and of all cereal urowths which incorporate lari>e quantities of silica in their conii»osition. This, anes have for us a hii»h geoloiiical interest and value. In the matter of Hawaiian soils we are first inb pressed with the low amount of the total siJioi and second- ly, with the hiiih proportion of this that is sidnble. Hoth these features distiugnish the soils of these islands from those of America, and of all ueoloiiically old countries, such as Euiiland and Germany, whence, thronuh the courtesies of Prof. Maercker, in (Tcrmany, and ^fessrs. Lawes and Gilbert, in England, we have received numerous analyses. ^lor(M)ver, the total silica found in Hawaiian soils is very much less than is found in the normal, solid lavas from which they were derived. In a com]>arison of the silica, we also repeat the statement of the iron and alumina in tlu^ lavas and soils. Materials. Si Oa Fe2 O3 Al, 0, Hawaiian Lavas Hawaiiau Soils Per cent. 47.590 27.927 Per cent. 15.02 36.44 Per cent. 19.92 22.63 105 It is thus seen that the soils contaiu 41.39;^, less silica than the lavas; and we have aln^ady seen that of the total silica still foiuid in the soils, 56.0% is "soluble silica." We have already said that these features distin guish Hawaiian soils from those of geologically older countries, and particularly those of America that we have submitted to an absolute analysis. Ileturning to American soils, it is not i)ossible, even with an approximation to accuracy, to submit these to a comparison witli the rocks froni which they have been derived, in the jjrecise mode followed with Hawaiian soils. There are regions in the United States where a true geological study of the soils may be made in con- nection with an examination of the rocks from which the soils actually came. And in the Western States, more especially where volcanic rocks (which Clarke and Hillebrand have shown to be normal basalts) have cov- ered notable areas, it may be possible to find relations of lavas and soils which, although now older, were primarily identical with the relations on these islands. We merely suggest that it may be ijossible. The American soils composing the collective sample that we are using in this comparison, it has already been said, are chiefly from regions where the rock formations are anterior to the carboniferous age, which is illustrated by a geological majj of the United States. These early formations fall under the grand divisions of Archaen, Silurian, and Devonian rocks, those of the latter two ages comprising the vaster portion of the areas represented. The Archaean formations include granitic, hornblendic, feldspathic and calcareous rocks, and also rocks rich in iron, and locally in titanic acid. The low^er and upper Silurians embrace sandstones, siliceous slates, claystf)nea 106 and shnlos; and also limestones on a larije scale. The lower Devonian rocks, according- to Dana, "represent the great limestone making- period of the age in America, whilst the later Devonian formations are mostly shalea or sandstones." These formations rnn on also into those of the snb-carboniferons period, during- which the initial work was begun in laying down the formations that mark the carboniferous age. In the selections that we have made from the superb series of analyses of American rocks made by Messrs. Clarke and Hillebrand, U. S. Geological Survey, which may well be used by agricultural chemists in America a» indicating the nature of a primary basis for an actually scientific study of soils in the United States, are found examples of the composition of most of the rocks which comprise the formations of the early geologic ages that have been named. The authors, however, would not claim that those analyses are adequate to enable us to estimate the mean composition of the rock masses which compose the surface formations covering the areas of the vast regions from which the soils were taken. The composition of the several kinds of rocks is furnished by these analyses; but the relative geographical areas that are occupied by the respective formations are not known, and without these anything of the nature of an approxi- mate average cannot be attained. Nevertheless, and while the average of the composition of the several kinds of American rocks will not furnish us with the knowledge that is rccjuired, the data can aid in our present pur])ose; so that we shall reproduce the mean of the compositions of the rocks by the side of the mineral constituents of the American soils . 107 Materials SiOz Fea AI2 Oe CaO MgO Ka Na^O American Rocks American Soils. Percent 53 93 80.54 Per cent 18.81 13.25 Per cent. 9.42 0.83 Per cent. 2.10 0.77 Per cent. 2 31 1.62 Per cent. 1.67 2.23 By this comparison we reacli results which are abso^ lutely the opposite of the results from comparison of the Hawaiian lavas and soils. The Hawaiian soils contain 27.92 parts less silica than the lavas; whilst the American soils contain 2G.G1 parts more silica than the rocks. Concerning the chief bases, the Hawaiian soils contain 24.13 parts more of iron and alumina than the lavas; and the American soils 5..56 parts less of these constit- uents than the rocks. In the matter of the silica, we again allude to the high content of soluble silica in Ha~ waiian, and the insoluble state of the silica in American soils. We have already said that the results of a comparison of American soils and rocks do not bear the same weight of significance as the comparison of Hawaiian soils and lavas, for the reason that we do not know the propor- tions of the several kinds of rocks that have gone to the forming of the American soils. We do not know that their averages of silica amounted to 53.93%. It may have been less, but probably was more. It appears quite certain, however, that the average of silica in the great rock masses did not contain anything like the amount of silica found in the American soils derived from those rocks. Even the granites contain only 72.0% of silica, or 8% less than the soils, and it appears further certain that the silica average could not be equal to that of the granites when the vast areas of claystones, slates, shales, loess, silicates and carbonates of iron, and limestones, 108 whose averages result iu 43.3% of silica, are iiu-liuled. Again, the iron and alumina found in the American soils indicate that formations of the nature of claystones, slates and shales have contributed notably to the produc- tion of the earths overlying thi^ vast areas traversed by glacial drift. Moreover, the rocks that would likely con- tribute the most to the formation of the drift soils wouhl be such as yielded more easily to the glacial action, and the siliceous sandstones, which contain less silica than similar residual products of rock disintegration on these Islands, and the granites would not be of that order. In brief, the indications are that those collective rock formations contained notably less silica than is found in the soils that have been derived from them, but we bear in mind that these are, at present, only indications. With these extraordinary marks of difference, in the structure and chemical composition, between Hawaiian and American soils the question naturally follows con^ cerning the causes and modes of change from which thia fundamental distinction has resulted. The American soils are old, and have, comparatively considered, reached a state beyond which little further change is possible. On the other hand, the soils of these Islands are all geologically very recent, and over certain areas they are very new, whilst special lava flows occurred so short a time ago that the rock cannot yet be said to be changed into soil. This youthful state of Hawaiian soils provides an excellent possibility of noting not onl}^ the causes by which the lavas have been resolved into earths, but also indications of the modes and slow processes by which soils of a given structural composition and type may undergo intrinsic change and pass over into soils of another type, having a totally different composition. 109 In the first place we shall repeat the comparison of Hawaiian lavas and soils. Materials. Si O2 Fe, 0, AI2O3 Ca MgO K2 NaaO Hawaiian Lavas Hawaiian Soils. Per cent. 47.90 27.92 Per cent. 13.36 36.44 Per cent 18.23 22.63 Per cent. 8.99 0.69 Per cent. 6.05 1.24 Per cent . 1.50 0.74 Per cent 2.20 1.42 This statement of lavas is based upon the inclusion of moisture and combined water; in other statements of the same lavas the water is excluded, in order to com- pare with other water-free materials. These data rep- resent the enormous change in the relative proportions of the elements in the course of passage from lava to soil. This change took place slowly, which is shown by the previous statement on the ^-weathering lavas;" and the soils are still in a state of change, which it is possible to illustrate. Of the total silica still in the soils, it has already been shown that 56% is "soluble silica,'- and liable to gradual removal. That it is being removed ia shown by an examination of the composition of the dis- charge waters leaving the four large islands, that is given in the second part of this work, in which it is seen that the silica contained in the waters is twice as great as either the lime or magnesia; five times greater than the potash; five times greater than the combined iron and alumina, and only equalled by the soda. Whether this soluble silica exists in the free state in the soils in part, or is wholly combined with the bases, has not yet been determined; but there are indications that the silicates of iron and alumina are slowly disintegrating, the silica 110 being carritMl away, and, especially the iron as gradually aggregating. In the matter of the removal of the bases it is seen that the lime has gone with the greatest facility — the element that is most vital when we come to the relations of the soil to plant life. Of the alkalies, soda has been the most resistant. It is remembered, however, that the most of the soda is bound up in the "insoluble residue," having resisted the solvent action of tlie concentrated hydrochloric acid. This behaviour of the soda, suggests a link of relationship betw^een the Hawaiian, and the old American soils. The alumina is slightly higher in the soils than in the lava; yet it has been removed on a vast scale. There w^ere five parts more of alumina than iron in the lavas; but there are fourteen parts more of iron than alumina in the soils. This denotes the relative behaviours of these bases in the changes that are trans- piring. In the course of examination of the different Hawaiian soils we have noted a difference, in degree, of behaviour of soils formed under dissimilar conditions. Soils deriv> ed from tufa lavas, which underwent chemical action during the initial stage of disintegration, contain ten parts less of total silica, and eight parts less of soluble silica, than the soils formed by the weathering of normal lavas in dry conditions. Again, the "upland" soils con- tain two parts less of total silica, and three parts less of soluble silica than the lowland soils. This results, in part, from the circumstance that tlie rainfall upon the up- lands is double that of the lowlands. It is unquestion- ably due also, to the further circumstance that the up- lands soils, with the large content of decaying organic Ill matter, ai:e five and one-half times more acid than the lowland soils, which we have carefully observed. The agricultural analyses have shown that the soils contain an amount of silica that was not disturbed by the digestion with strong hj'drochloric acid, and not removable from the insoluble residue with alkalies; which is understood as "insoluble silica." In the course of the advancing disintegration of the soils, and of the separation and removal of the soluble silica, the stage in the history must be reached when the insoluble silica must become an increasing factor, and not only in rela- tion to the soluble silica, but as a constituent of the whole soil. For it is not only shown that the soluble silica is being actually carried aw^ay, there are numerous indications that even the iron, which has accumulated so enormously in passing over from the lavas into the soils, reaches a degree of concentration when it separates from the other constituents, and forms layers and con- cretions of iron ore. We have, in a former table, given four analyses of such concretions, where the separation of the iron had been jjrecipitated by chemical action on the lavas, the average of which gave 78.41% of iron oxide. These concretions we have found on an appre- ciable scale on the older Islands of Oahu and Kauai. In relation to the stage or time when the "insoluble silica" shall begin to assert itself in the composition of the soils, it is noted that the upland soils already, not only contain more insoluble than soluble silica, but also more insoluble silica than the lowland soils. It is here suggested that rainfall and acidity are operating along the line of a change in the type of the soils. In the con- ditions of nature the change must be very slow. In another place we have shown that when an one per 112 cent, solution of citric acid had acted on a soft lava for about fifty days, the lava Avas wholly taken to pieces, the bases being chiefly in solution, leaving a residue com- posed of insoluble silica and silicates. This disintegra- tion with the weak organic acid required fifty days for completion. Proceeding from the weak, to the use of a strong acid, it is shown by the agricultural analyses that the same was accomplished on the soils in ten hours, and leaving insoluble residiies containing over 75% of insoluble silica and titanic acid. These insoluble residues however, comprise only 31.1% of the original soil; so that to produce the insoluble residues of their present compo- sition, 68.9%) of the more soluble parts of the soil had to be removed. We cannot compute how long a time might be required by Nature to accomplish what was done by the strong hydrochloric acid in ten hours. More- over, we do not know that the sum of natural processes will be exactly in the same direction, or that they will lead to the same results. At this place, however, we call attention to the impressive circumstance that when Ha- waiian soils are digested with concentrated hydrochloric acid, as we have already explained, an "insoluble residue" remains which is almost identical in chemical composi- tion with the American soils. We bring these into com- parison; — the insoluble rc^sidues, all of which are pre- served, representing some 1300 sub-samples of Hawaiian soils. Materials. SiO, Fe-AKOe CaO Mg K.O Per cent 0.70 1.62 Na„0 " Insolulile Kesidues, " Hawaiian Soils American Soils Per cent 76 53 80.54 Per cent. 18 66 13 25 Per cent. 0.71 1 83 Per cent. 0.86 1.77 Per cent. 1.64 2.23 113 It is seen that by acting- upon Hawaiian soils with strong hydrochloric acid for ten hours, and removing the acid soluble constituents, we have an insoluble residue left, of the same color and similar composition as the American soil, according to the absolute analysis of the latter. If the hj^drochloric acid had removed 5% more of the alumina and iron, the silica in the insoluble residue would have been exactly equal to the amount in the American soils. We have previously said, however, that a further treatment of the insoluble residue with a fresh quantity of the strong acid did not appreciably reduce its weight or composition. We have also said that we do not know that the natural processes are acting exactly along the same line, or will lead to the same results as are accomplished by the strong hydrochloric acid in ten hours. We have however, furnished such natural indications as we have observed, showing that the slow action of dilute acids, or more properly speaking of acidulated waters, does appear to be moving towards the results which are more instantly produced by the strong acid. If these results should be eventually reached, and the soils of these Islands be converted into a type represented by the insoluble residue, and resem- bling American soils, the time required to that end will be immense; since, as it has been said, no less than 69% of the present, more soluble constituents have to be separated and removed. These considerations bearing on Hawaiian soils lead to the question concerning the mode and processes by which the American type of drift soils, that we have examined, arrived at their jiresent composition? Reasons have already been given for deciding that those American 114 soils have been derived from rock formations that must, in the agj»re<>ate, have contained a much h)wer content of silica than is found in th(- soils. In fact there is not any order of rocks but the siliceous sandstones that con« tains anythin<>- like so much silica as the drift soils that we are speaking of. Those sandstones however, are essentially separation or residual products of previous formations; resembling the residual siliceous productft from the disintegration of Hawaiian lavas, that are to form future sandstones, and in which our analysis found 93% of silica, or 5% more than the mean of American sandstones. If then the rocks, and also the soils during an earlier period in their history, contained less silica than is now in the present soils, they must necessarily have contained more bases, and the question has already been put concerning the j^rocesses by which the soluble constituents have been removed, and the soils have reach- ed their present composition, with the enormous content of silica. We have been specially concerned with the mode of formation of Hawaiian soils, and only gen- erally interested, and for the value of comparison, with American soils. It has appeared to us, how- ever, that such American soils as we have examined have reached their present composition and character by a long history of change that has proceeded by a course of processes similar to what has been described. We have endeavored to observe some remaining records of such processes in soils from special localities. By the courtesy of Professor Goodell, of the Amherst Experi- ment Station, Mass., we received analyses of soils in the Connecticut valley, lying at the feet of Mount Holyoke and Mount Tom, that are amongst the grandest results 115 of the great eruptions of the Triassic period, and which are composed of basalts that, according to Dana, have the same composition as the lavas of the Hawaiian Islands. Those soils, however, show an average "insolu- ble residue" of 85.37%, according to the agricultural analyses; which indicates, as we thought probable, that the Connecticut valley soils are soils of deposit, and bear no necessary relation to the rocks of the immediate local- ity; or that they have become so radically altered in type as not to resemble, in the least, the basaltic soils of these islands. Our inquiries concerning areas in Europe, with which we are also familiar, have been just as resultless in this respect. Yet we are impressed that there are localities in the United States where such records may be found in the soils. Were this the place for such a dis- cussion, it might be possible to present the most impres- sive indications showing that not only have the older soils been brought to their present state of composition by the processes recounted, but that those changes trans- piring in the structure and composition of soils between the periods of virginity and old age are only a part and continuance of the initial processes of disintegration whereby the rocks were resolved into earths. Further, that it has been by the removal of the soluble constit- uents from soils, resulting in the changes of type, as well as from the initial decomposition of rocks, that the materials have been derived for the laying down of the grand and successive series of formations which form the superficial crust of the earth. How superficial the crust is that bias undergone con« tinual dissolution and deposition in new formations it is not possible to say. Of possible interest at this place, 116 it is remarli^ed that the specific gravitj' of Hawaiian soils has been found to be 2.87, as compared with about 2.6, tlie specific gravity of certain American soils. The specific gravity of the Hawaiian soils, free from com- bustible matter, is approximately 3.4; compared with the lavas, having a specific gravity of almost exactly 8.0. It is seen that the specific gravity of the lavas is very little greater than that of the natural soils, or than the moan specific gravity of tlie surface of the crust of the earth. The specific gravity of the earth as a body, however, is .5.5. If then, we compare what must be the relatively smaller mass of the interior of the earth, hav- ing a density so great as to bring the specific gravity of the globe up to 5.5, with the greater mass of the exterior of low density, it then appears that our lavas must come from a depth merely beneath the surface, comparatively speaking; and that the locality of their origin may bear no relation to the more profound internal depths and con ditions of our globe. With such questions, liowever, we have nothing to do at this place and time. Our considerations, so far, have dealt only with the origin and nature of Hawaiian soils. This knowledge is preliminary' to any further investigations. We are, however, practically concerned with the economic rela tion of the soils to plant life and in'oduclion, and this re- quires a knowledge, not only of the elements of plant food contained in the soils, but also of their condition of fitness for use by the growing crops. Consequently we have had to look into the state of arailahUiti/ of the essen- tial constituents; the results of which furtlier investiga- tions will be given in the sccoihI pdrt of this work. AVAILABILITY AND LOSS OF THE ELEMENTS OF PLANT FOOD IN HAWAIIAN SOILS. In the First Part of these investigations attention was confined to the "Origin and Nature of Hawaiian Soils." We shall now endeavor to obtain some more definite and special knowledge of the solubility of these soils, and try to understand something of the state of avail- ability of certain required elements in plant growth, and their behaviour under the action of processes operat- ing in Nature. Before proceeding, it may be well to repeat how strict- ly necessary it is that the fundamental differences which we have shown do actually obtain between Hawaiian soils, and soils of America and other countries, in the matters of relative and structural composition, and in the state of solubility, should be continually kept in mind. These differences persuade us at the outset that corresponding results cannot follow the action of either artificial solvents, or the processes of Nature, upon soils 118 of suL'li dissimilar composition. The primary circum- stance, that we are dealino- with soils of a strongly basic nature, as compared with other soils of a highly acid character, suggests that the action of any solvents may be widely different upon the two orders of soils. These considerations then, should guard against the use of re~ suits found upon Hawaiian soils, and in these conditions, in judging of the character of other and different soils, found in totally different conditions. LABORATORY MODES OF ESTIMATING THE ELEMENTS OF PLANT FOOD AVAIL- ABLE IN SOILS. The annals of agricultural chemistry furnish the re- sults of numerous and very dissimilar endeavors to establish a means of estimating the proportion of the elements important in plant nutrition which may be said to be availahic at the time of examination, for that pur- pose. Up to this time, those endeavors have not resulted in any methods which appear to be in so far uniform as to agree in the recognition of the principles upon which, it may be found, such methods must rest. The most representative associations of agricultural chemists continue to oscillate between extremes that indicate the absence of anything permanent in inMnciple, or that prom- ises to become uniform in practice. Methods and Solvents. — In framing a method, and in the selection of solvents, for estimating the proportion of plant food possibly available in soils at the time of ex- amination it seems necessary to be guided by an exact observance of the agencies by means of which the insol- 119 iible soil-materials are being- daily altered by the pro- cesses of Nature in the field into forms in which they can be used by growing plants. The processes operating in Nature by which the elements are prepared as food, are physio-chemical; and for this reason the problem cannot be primarily con- sidered from the analytical standpoint The solvent agents operating in Nature, in addition to ivater, are the acids moving in the sap of living plants, and operating on soils through the membranes of their roots, the chief one, as far as we know at present, being carbonic acid (COo); and, more important, the acids which result from the decaj' of vegetable matter upon and within the soil. The acids formed when plants, roots, and fruits decay are simple organic acids — carbon acids; and the amido acids — carbo-nitrogen acids. Therefore the acids in liv- ing, and produced by dying plant organisms are carbon acids, with or without nitrogen. In the complete decay of vegetable matter these organic acids are resolved into ultimate mineral bodies; the carbon into carbonic acid, and the nitrogen into nitric acid, or nitrogen, the simple forms in which these were primarily taken from the air to build up the plant organism. Consequently the amounts of carbon and of nitrogen contained in plant organisms are respectively the measure of the relative amounts of simple carbon acids, and of amido acids that can be produced in vegetable decay, and of the amounts of carbonic and nitric acids that finally result from that decay, and which act as solvent agents on the soil. The minute amount of sulphuric acid, and the still smaller l^ortion of phosphoric acid, that are formed from tlie 120 suli)hui' ill tlie pi'oteids and iiucleiiis, and from the I)li()spliorous in the phospho-glyeerides (lecithines) are unnoticed now. In the absence of elementary estimations of the carbon and nitrogen in plant oroanisms, these estimations being- confined to constitnent bodies, we may come a])proxi- mately to such determinations by ascertaining the amount of tlie constituents of plants that are composed of carbonaceous bodies not containing nitrogen, and the proportion of bodies that do contain nitrogen. The bodies free from nitrogen are the so-called nitrogen-free extract matter, the fiber, and, for our present purpose are added the fats. The bodies containing nitrogen are now col- lectively considered as proteids. The amounts of these nitrogenous and non-nitrogenous carbonaceous bodies found in a broadly representative series of agricultural plants are found in the following table: Materials. No of Examples Proteids. Fiber. Nitrogen-free Extract-Matter Fats. Legumes and Cereals Root aud Biilbs Graiu and other Seeds 32 14 45 Per cent. 8.0 13 9 12 9 Per cent. 27.6 10 5 2.3 Per cent. 51.5 64 5 79.5 Per cent. 3.1 3 1 4 4 Means 91 11.6 13.5 65 2 3.5 If the third series, tlie seeds, are excluded for the reason that the grain and seeds are not allowed to return direct- ly to the soil, the means will remain nearly the same, the large ])r<>portion of crtriicl iiiafh rs in seeds being off- set by the small amount of fibei'. These data show that in the 91 examples of growths we have. Nitrogen-free Carbnnaeeous bodie.s = 82.2 Per cent. Nitrosenons C ;rbonaceons bodies =116 "' 121 The nitrogen-free bodies may be considered as con- taining six parts of carbon (C,j Hio O.5). The proteids, in which the elementary analysis finds 16% of nitrogen, with 54% of carbon, are bodies in which, according to the relative atomic weights, abont three parts of carbon are associated with one part of nitrogen. The relation of the carbon and nitrogen present in these organisms, then may be expressed thus: Per cent. r. F'arts of Nitrogen-free Carbonaceous bodies.-. . Nitrogenous Carbonaceous bodies Nitrogenous Carbonaceous bodies 82.2 X 11.6 X 11.6 X 6 = 3 = N. 1 = 493.2 Carbon 34.8 528 11.6 Nitrogen These data indicate that in the composition of the plants, roots, and seeds stated there are forty-five parts of carbon to one part of nitrogen. Therefore in the de^ composition of those organisms there must finally be pro- duced forty-five parts of carbonic acid and one part of nitric acid. Mtric acid is a more immediately active solvent than carbonic acid, and will dissolve soil material rapidly while its action lasts. The duration and measure of its action, however, are fixed by the quantity, and can ex- tend only to the point of neutralization with the bases it acts upon, which is the same with the carbonic acid. Moreover, nitric acid is a mono-basic acid, while carbonic acid is di-basic; which thus doubles the solvent power of the forty-five parts of carbon and lowers the possible action of the one part of nitrogen to only one-ninetieth (1-90) part of that of carbonic acid, providing both acids exercise their action on the soil bases to neutralization. 122 The above considerations have appeared to the writer to be a guide in the selection of solvents which are to compare, in any measure, wdth tlu^ action of ])rocesses operating in the tield; and tlicy led to the exclusion of mineral acids, and the use of simple carbon acids and amido acids in these investigations. Citric Acid as a Solvent. — It is due to consider, the first in order, citric acid as a solvent, since not only has considerable work been done, and conclusions been reached, based u])on the solvent action of this acid, it also exercises an accepted infl-uence in the estimation of certain fertilizing materials of great economic impor- tance. As an exception to omitting references to author- ities generally, which is done to economize space, atten- tion is called to the work of Dr. Bernard Dyer with citric acid, on soils. The examinations and results to be recorded are not confined to observations made upon samples of soil taken by chance, but upon soils definitely selected in districts of the four islands. On Oaliu, samples were taken from Waianae, P^wa, Heeia and Waimanalo districts; on Kauai, from the Kekaha, Makaweli, Kealia and Kilauea districts; on Maui, from Hana, Paia and Wailuku dis- tricts; on Hawaii, from Hawi, Niulii, Ookala, Laupahoe- hoe, Hilo and Kau districts. These comprise 30, what we shall call, ti/pc samples: and 3G0 siih-samples : which means that each type sample represents some 10 to 15 sub-sam- ples taken from a large area where the soil is of one type. As we have already remarked, the samples were taken personally by the Avriter, or in places, and by methods, advised by us. The superficial character of the land, eh'vation, exposure, and rainfall were noted, and attached to each sample, witli its agricultural analysis. 123 In addition to the care given to make the soils selected representative of the types found on all the islands, these soils were resolved into two groups: Upland (mauka) soils; and lowland (makai) soils. This division was made in our rejDort on soils in 1895, and has been con- tinued since, and denotes differences in the soils, which are mainly due to rainfall and other superficial causes. In the thirty (30) type samples of soils spoken of, estimations were made: First: Of the lime, potash, and phosphoric acid soluble in water. Second : Of the lime, potash and phosphoric acid solu^ ble in an 1% solution of citric acid. At this place we state simply the mode of treat- ing the soils with the solvents, leaving some observations to be made on methods for a later occasion. In the estimation of elements soluble in water 200 grams of the field sample (not fine earth) were put into a closed funnel, with a ground glass cover, and treated with lOOOcc of water for 48 hours, the water percolating through slowly, and then returned upon the soil, and continued for the time stated. In determining the elements soluble in 1% citric acid solution, 200 grams of the field sample were put into a two-liter bottle with 1000 cc of the 1% citric acid solu- tion; the bottle was gently shaken every fifteen minutes during the day portion of 24 hours, and at the end of this time filtered off. One reason for the shorter time of treat- ment with citric acid is its liability to fermentation in great dilution, a matter that will be spoken of later. With these brief definitions, we now give tables of data which set forth the solvent action of water and 124 citric acid ros])ectivelT upon each of 30 type samples of soil, taken fi-om districts on the four Islands, and rep- resentative of all our soils. We give the amounts of lime, potash, and phosphoric acid found by the agricul- tural analysis in each soil, with the amounts soluble in water and citric acid respectiyely, and the pounds per acre actually dissolved by the solvents. The individual statement of the behaviour of each soil under the action of the solvents is necessary in order to understand the range of variation, and the different forms of combina- tion in which the elements exist in the soil, which is in- dicated by the variable ratio of solubility. The data occupy much space; but they also involved great labor and time in their preparation. We give first the data on lAme. 125 LIME. Laboratory Lime in Soluble in Pounds Soluble in Pounds per No. Soil. Water. per Acre. Citric Acid. Acre. Per cent. Per cent. Per cent. 3 0.257 0.0057 199 0.145 5075 5 0.112 0.0035 122 0.075 ^625 6 0.442 C 0029 101 0.195 6825 12 0.330 0.0087 304 0.097 3395 13 0.600 0.053 18.55 6195 14 0.510 '" 0.6682'" ' ' 287 ' ' 0.177 17 0.435 0.2?8 7980 3605 19 0.338 " 0.6665"" ' ' 207 " ' 0.103 22 0.200 O.Ofil 2135 25 0.800 ""o'.6ii6"' ■ ■ 406 ' ' 0.233 81.55 26 0.215 0.0024 84 0.043 1505 43 0.448 0.0102 357 0.230 8050 45 0.541 0.147 5135 46 0.473 0.118 4130 47 0.168 " 0.0041 " "Hi" 0.113 3955 55 0.365 0.0103 361 0.177 6195 60 0.080 0.0018 63 0.055 1925 33 0.320 0.099 3465 1050 66 0.109 ""o'.6669"' ""si" 0!030 69 0.350 0.0024 84 0.155 5422 71 0.448 0.0046 161 0.171 5985 84 0.095 0.0014 49 034 1190 90 0.120 0.0009 31 0.043 1505 92 0.157 0.0018 63 0.031 1085 93 0.092 0.0080 280 0.068 2380 601 0943 0.0026 91 0.281 9835 604 0.473 0.0035 122 0.087 3045 614 0.237 0.059 2065 616 0529 0.305 617 0.025 875 126 POTASH. Laboratory Potash in Soluble in Pounds Soluble in Pounds No. Soil. Water. Per Acre. Citric Acid. Per Acre . Per cent. Per cent. Per cent. Per cent. 3 0.250 • 0.0070 245 0.048 1680 5 0.595 0.0025 87 0.023 805 6 0.201 0.0016 56 0.009 315 12 0.178 0.0029 106 0.025 875 13 0.250 0.299 0.055 0.084 1925 14 ""o.oiis" ""508"' 2940 17 0.405 0.257 0.047 0.068 1645 19 '"00236" ""826 ' 2380 22 0.221 0.435 0.014 0.053 490 25 0.6687 ' ' ' '304 ' 1855 26 0.396 0.0029 102 0.027 945 43 0.129 0.0025 87 0.052 1820 45 0.139 0.250 0.259 0.030 0.039 0.017 1050 46 1365 47 "0.0625"' • • • • y ^ • • • 595 55 507 60 0,291 ""o'.6o'2o"' ■ " 70"' 0.015 525 "" 33 0.280 0.0021 73 0.062 2170 66 0.234 0.0082 112 0.043 1505 09 0.426 0.0038 133 0.039 1365 71 0.413 0.0059 206 0.051 1785 84 0.243 0.0020 70 0.013 455 90 0.407 0.0032 112 0.014 490 92 0.315 0.0063 220 022 770 93 0.242 0.0053 186 0.013 455 601 0.400 0.0038 133 0.021 735 604 0.166 o.o( ).")(; 196 016 560 614 0.422 0.0033 115 0.025 875 616 0.258 0.349 617 127 PHOSPHORIC ACID. Laboratory Phosphoric Soluble in Pounds So uble in Pounds per No. Acid in Soil. Water. per Acre. Citric Acid Acre. Per cent. Per cent Per cent. 3 0.211 trace trace 0.0043 150 5 0.173 trace trace 0.0035 122 6 0.365 0.0006 21 0.0125 437 12 0.096 0.0002 7 0.0032 112 13 0.080 0.0002 7 0.0105 367 14 0.115 trace trace 0.0012 42 17 0.157 trace trace 0.0042 147 19 0.166 0.0002 7 0.0018 63 22 0.223 0.221 0.0020 0.0019 70 25 trace trace 66 26 0.153 trMce trace 0.0018 63 43 0.110 trace trace 0.0071 248 45 0.085 0.088 0.292 0.0038 0.0036 0.0022 133 46 126 47 trace trace 77 55 0.464 0.354 0.0035 0.0016 123 60 trace trace 56 33 0.137 0.0008 28 0.0034 119 66 0.579 0006 21 0.0039 136 69 0.602 0.0002 7 0.0058 203 71 0.307 0.0002 7 0.0027 95 84 0.527 0.0002 7 0.0041 144 90 0.606 0.0005 17 0.0031 108 92 0.281 0.0005 17 0.0025 88 93 0.772 0.0002 7 0.0057 199 601 0.711 0.0007 ; 24 0084 294 604 0.301 0.0003 ! 10 0.0085 297 614 0.239 0.0005 1 1^ 0.0038 133 616 0.132 0.333 1 0.0033 0.0046 115 617 161 128 It may be explaiiiod at this place that a statement iu per cent, quantity up to the fourth decimal may be accept- ed as reliable, for the reason that in all the determina- tions never less than 200 grams of soil were taken. Con« sequently an amount of phosphoric acid which is express- ed by even a low figure in the fourth decimal means an actual weight of magnesium pyrophosphate that a moderately sensitive balance will easily take account of, even up to one-ten-thousandths of one per cent. From the tables of data presented we obtain a more or less adequate understanding of the variation in the amounts of lime, potash, and phosphoric acid present in our soils, and of the different behaviour of these elements, under the solvent action of water and citric acid, in soils from different localities. These differences indicate that these elements are contained in the soils in varying chemical forms, to which can be due the difference in solubility. Before making further comment u])on the tables of data we shall bring them together in concise averages; and at the same time we shall divide the soils examined into the two classes under which they have been considered previously, viz: Upland (mauka) soils, and lowland (makai) soils, this division being based uiK)n the signal difference in climatic conditions that obtain at different elevations. The first small table gives the aver- age of lime, potash, and phosphoric acid in the 30 type soils, as shoAvn by the agricultural analyses: Soils. Type Samples. Sub- Samples. Lime. Potash. Phosphoric Acid. Uplands . 14 16 168 192 Per cent 277 0.399 Per cent. 0.299 0.377 Per cent. 0.316 Lowlands 0.304 129 The elements soluble in water, and in one per cent, citric acid solution, are respectively as follows: Soluble in Water. Pounds Per Acre. Soluble in Citric Acid. Pounds Per Acre. Total Pounds Per Acre. Uplands. Lime Per cent. 0.0032 0.0031 0.0001 0.0054 0.0047 0.0003 112 110 3 189 164 10 Per cent. 0.0940 0.0250 0.0035 01330 0.C380 0.0046 3290 875 122 4655 1330 162 9695 Potash Phosphoric Acid .... Lowlands. Lime Potash 10465 11795 13965 11795 Phosphoric Acid .... 10640 The "pounds -pev acre" in the three columns are calcu- lated by means of the weight of an acre of soil one foot deep, to which usually the sample is taken, which weight is set at the average of 3,500,000 lbs. The weight is greater than this on most lowJaiids and less on the up- lauds, which contain notably more organic matter. The weight of the cubic foot of soil varies some 13 lbs. — from 80 lbs. to 93 lbs. for water-free material. These estimations of solubility of the lime, potash, and phosphoric acid confirm in an ample measure our obser- vations made upon upland and lowland soils in 1895 and 1896. In speaking of the lime contents of the upper and lower lands it was said in 1895 "the lime present in the upland soils has been, previous to cultivation, prac- tically the same, which is shown by analyses of virgin and cropped soils, and comparison of the upland virgin soils with the lowland soils;" but "as the lime, being in a more or less soluble state, was w^ashed out and down into the lowlands, or into the sea, by the heavy rains." The lime soluble in water, we see, is 189 lbs. in the low- 9 130 lands, against 112 lbs. in the nplands. The same is found under the action of the citric acid, where the upland soils give up 3290 lbs. per acre of lime, and the lowlands no less than 4655 lbs. per acre, in 24 hours to the action of citric acid. The observations on the solubility of the potash in the upper and lower lands are also in agree- ment with i)revious findings. In the matter of the phos- phoric acid the results are even more remarkable in the way in which they suj)port previous conclusions upon this body and its state of insolubility. In our reports of 1895 and 1896 it is said "the phosphoric acid in our soils ap- pears to be locked up so securely by the iron and other compounds that plants use it with difificulty." In the above tables it is seen that the phosphoric acid soluble in water in the lowlands is 10 lbs. per acre, and in the uplands 3 lbs., which, practically speaking, means none at all, whilst the amount dissolved by the citric acid is relatively very small. By far, the most striking result from the action of an 1% solution of citric acid upon the soil is the enormous proportion of the lime, and also of the potash, dissolved. The lime found by the agricultural analyses in the ui)land soils was 9695 lbs. per acre, and of this amount 3290 lbs., or 30%, were dissolved by the acid in 24 hours. In the lowland soils the lime ]ier acre was 13,965 lbs., and the citric acid took out 4655 lbs., or about 30% of the total found in the soil by the agricultural analyses. The potash dissolved was also a very large proportion of the whole. Yet the action of an 1% citric acid solution is much less than the action of solvents that are used by chemists in estimating the so-called "amount of available elements." The question suggested by our results with citric acid, 131 and more- strongly reiterated by other results that have been published, is How much nearer are we brought to an understanding of the actual availability of the elements — their relation to the demands of the plant — by the use of citric acid in 1% solution, as compared with the results of the agricultural analyses obtained by the use of con^ centrated hydrochloric acid? The latter mineral acid^ in 10 hours, upon the water bath, dissolved out all the lime which, it has been said, "can ever become soluble;'^ and yet a cold, 1% solution of citric acid, which is the solvent most accepted for determining the amount of the elements "immediately available," takes out 30% of the "total lime" in 24 hours. The "absolute analyses'' of soils given in the first part of these investigations, with the examination of the "in- soluble residue" of the soil, after digestion with con- centrated hydrochloric acid, afford notable light on the question of solubility of the constituent compounds of soils. In the "insoluble residue" of Hawaiian soils we found 1.02% of lime, which was left after the action of concentrated hydrochloric acid, and the repeated action with a new quantity of the acid. Consequently the "absolute analysis" of the soil found O.Tir)% of lime, in- stead of 0.342%, recorded by the usual aqrk'uJtural mmhjsis. It is thus seen that the hydrochloric acid doe» not take out quite one-half of the lime present in the soil. These data afford, as said, notable light upon the state of solubility of our soils, and upon their mode of structure, all of which indications are invaluable in a complete study, and accentuate the fundamental prin- ciple of comparative solKhilitics which dominates all struc- ture in the mineral, vegetable, and animal kingdoms. The data, however, do not help us in estimating the lime, 132 potash and [►li<>si>li(>i'ic acid "immediately available'' to the growing plant. Knowing the toftil rune (tctiidllii contained in the soil, as shown by the "absolute analysis," and the pro])ortion of that total found soluble in concentrated hydrochloric acid aftei* digesting 10 hours upon a water bath, we went on to the examination with an 1% solution of citric acid. 'Phe citric acid solution, however, exercised a sol- vent action in 21 hours equal to one-third of the solvent power of the hydrochloric acid, dissolving out one-seventh of the total lime found by absolute analysis, and in the short period of one day. In this action we do not find anything, in point of degree, corresponding to the pro- cesses which go on in Nature; or in the results of this ac- tion, any indication of the proportion of plant food that is actually available. But for the grand principle of differential solubility obtaining in all matter, the citric acid would have taken all the lime out of the soil in probably less than six days. There is no relation between these results and the amount of plant food removed from the soil hj crops — which action of crops, however, as we shall show later, also bears no constant relation to the ratio of plant food depletion in soils under cultiva- tion, and therefore cannot be taken as a guide in the mat- ter of fertilization, or restoring the lost elements. Of course, it has not been claimed that the action of an 1% solution of citric acid corresponds to the action of solvent agents operating in Nature, but the results of its action have been taken to furnish an estimate of the available ]dant food in soils, and we find that it has but little more meaning in sudi respect than the hydrochloric acid used in the common analyses. What the continued action of an 1% solution of citric 133 acid upon soils might result in is exemplified by results obtained b}^ us of its action on lavas. A weighed piece of solid lava was put into a bottle, and 300 cc. of an 1% solution of citric acid added, the acid being renewed every fourth day on account of its liability to ferment. Another weighed piece of the same lava, but in a state of decomposition, from simple weathering, was put into a second bottle and treated in the same way with the citric acid. At the end of three months the weathered lava was totally disintegrated, and chiefly in solution. The solid lava, which weighed 29.140 grams on Jan. 27, after nine months action of the citric acid, weighed only 14.63 grams, the residue being left in such a soft state that the thumb nail penetrated it easily. In considering a solvent, however, which is selected with a view to approximating the action of the solvent agents operating in Nature, some attention should be given to the organisms of plants, and their sensibility to the action of acids. An extensive series of observa- tions upon the "Relative Sensibility of Plants to Acidity in Soils" was conducted by the writer which furnished data, a part of which will serve the present purpose. The following data set forth the effect of applying to plants growing in tubs a volume of water equal to the amount that the soil w^as just capable of absorbing and holding, which amount was 48% on the weight of the soil. This water contained one-tenth of one per cent, of citric acid. Every fourth day the weights of the tubs were taken, and the amount of evaporation found, when enough water was again added to bring up the volume to the nmximum absorptive power of the soil. With the water added, enough citric acid in solution was also added to make the total water in the soil every fourth day equal to one- 134 tenth of one per cent, solution of the acid. The acid was applied with the greatest care by use of a large pipette, thus making sure that the same volume was delivered to each of the 18 plants under treatment. We give the results obtained by the use of one strength of acid which will be enough for the present purpose. Observations with different strengths of acid were made, the results of all which have already been published. The following data are the record of the action of a solution of citric acid, one-tenth of one per cent, in strength: A. CRUCIFEKA. Name of Plants Planted. Up. Failed Developement. Black Mustard. . White Mustard.. Sn- tliia course of time. So:\rE Prelimixary Tests. — In considerinli=10.2ec. alkali. In the first test the result of passing the solution throu<»h the same soil was to increase its lime, to double its potash, and to reduce its phosphoric acid contents. This appears remarkable (althouii,h we bear in mind the basic nature of the soil); for it is seen that the soil took up more phosphoric acid of which it already possessed an enormous quantity, and gave up to the passing solu- tion a double quantity of potash, although the potash in soil A is only two-thirds of the quantity in soil B, In the S(TO)i(l test, it is seen that, despite the very high potash content of soil B, that soil took one-half of the potash out of the solution on its passing through; also the same soil, although its content of phosphoric acid was only one-fourth as large as that of soil A, did not absorb any phosphoric acid from the passing solution. These, and other similar observations have led us to note that there is not any necessary relation between the amount of an element already contained by a soil and the amount that the soil will absorb. It is indicated that the power to ahsorh is controlled less by the quantity that the soil contains, and decidedly more by the chemical form in which it is contained, that has been hitherto understood. The behavior of acid and neutral soils in relation to ab- 159 sorption caused us to thiuk that the acid or neutral re- action of the solution would effect the absorptive power of the soil through which it was passed; and, consequent- 1} , that the power of soils in the field to hold back and prevent the loss of elements bought into solution by water may be partly controlled by their acid or neutral character. Observations bearing on this question were made upon a mixed soil, which was made up of equal quantities taken from 100 sub-samples. This mixed soil was treated with one per cent, citric acid for 24 hours, when the solution was separated from the soil by filtra- tion. This solution thus obtained, and containing the elements dissolved, was used for the tests, which were made as follows — 200 grams of the said "mixed soil" were put into a funnel, with a ground top and glass cover plate; a given volume of the said solution was poured upon the soil in the funnel, and allowed to pass through slowly, this being controlled by a piece of rubber tube upon the stem of the funnel, on which was put a pinch- cock. The solution was passed through at the rate of 30 drops per minute, and when run through, it was re-pass- ed through, thus continuing for 72 hours in each test. In explanation of the results to be given — A — gives the elements in the solution, which was still acid. B — gives the elements in the solution after been passed through a fresh quantity of soil. C — gives the elements in the solution, which was neu- tralized with carbonate of soda, before being passed through the fresh soil. D — gives the elements in the solution which was pass- ed through the original soil from which the solution was IGO obtained, and at the same rate, and for tlie same length of time (72 honrs) as in tests B and 0. Solutions. Original Solution Solution after passing througli fresh Soil Solution neutralized, and passed through fresh Soil . Solution after passing through the Original Soil A B D Ca O Per cent 0.1387 0.1629 0.0633 0.1535 K„ o Per cent. Ps Os Per cent. Fe» AI2 Oe Si Oj Per cent Per cent. 0.0228 0.0052 1.0271 !0.2010 0350:0 0028 0.8938 10.1040 0214 0.0031 6414 |o,0591 0408 0050 1.3130 0.2258 Test D — shows that the resnlt of continning to pass the solution through the soil from which it had be(^n obtained was merely to dissolve out more of the several elements, excepting phosphoric acid. Test C — shows an emphatic absorption by the fresh soil of all elements, notably of the lime, silicic acid, and iron. This is in particular agreement with observations upon the action of dilute citric acid upon neutral soils in dis- tinction from the action of the same solvent on acid soils. Test B — shows that in the absence of the carbonic acid, furnished by the carbonate of soda, as in test C, the lime and also the ])()tasli, continue to increase in the solution when it is passed through the frcsli mil. But almost one-half of the silica and phosphoric acid are taken out of the solution by the fi*(*sh soil, with a notable amount of the iron and alumina bases. Certain of our obsers^atious are in agreement witli, and others are emphatically opposed to, the findings of distinguished chemists, from the researches of Way down to more recent date. This, however, is more to be expected than wondered at. The absorption experiments conducted bv all other chemists were made with soils 161 either of a moderately, or highly acidic character; whilst these investigations are with soils of an ultra basic nature and this fundamental difference in the soils must exercise a determining effect in their relation to solvents used in ascertaining the so-called availability of their constituents, and in their property of absorption. The results that we have given, with other obseiwa- tions not recorded, indicate that, in the matter of Ha- waiian soils, the action on the one hand, of solvents upon soils, and on the other hand, of soils upon solutions, is controlled by the following factors: 1. The basic, or non-acidic nature of the soils. 2. Their structural composition, or difference of chemical form in which the constituent elements are present. 3. The neutral or acid reaction of the soils, due to free organic acids derived from the decay of less or greater amounts of vegetable matter, as found in upland and low- land soils 4. As affecting absorption, the kind of acid in the solu- tion, which will be shown at a later place. 5. The acid, neutral, or alkaline reaction of the so- lution containing the elements, and that is to be passed through, or brought in contact with the soils. Our several lahoratory modes of trying to estimate the plant food availahle in soils, which have caused us to traverse a very extensive ground of observation, have furnished knowledge of great value, and having a far- reaching interest. The results, however, have not pro- vided the precise information that we sought, but have indicated that that information is not to be found along lines, and by methods, of pure artificial research, and that 1G2 we iiiust ji;o out into Nature, aud uote the results of her processes in the field. Having done this, it may then be found possible to use the findings in the field, in combina- tion with adjusted methods in the laboratory, to guide us more actually in dealing with questions that come up in ever^^day practice. ELEINIENTS OF TLANT FOOD REMOVED FROM HAWAIIAN SOILS BY WATER AND CROPPING. In an early paragraph of this part of our investigations we endeavored to indicate the materials in soils which provide the solvent agents which operate through natural processes in rendering the insoluble elements suitable for plant food. We dwelt upon the decay of vegetable matter, and the acids that result from the decay, which, with water, carry on the work of food-preparation. We did not attempt, however, to follow the minute and com- plicated processes which Nature, with an unknown diversity of detail, may pursue in working out her ends. These may be infinitely intricate, or they may be more simple than we at present can grasp. If we cannot follow her methods, we can judge of their results; and these are recorded, on a grand scale, in cer- tain of the final processes by which her work is carried on and matured. Elements Removed by Waters of Discharge. — In the endeavor to find a brief and comprehensive expression of the results of the enormous and manifold action of water, and its leaching power on soils, we are usually referred to the composition of sea loater. So eminent an authority as Professor Hilgard errs in this matter, and says "the usual nature of the substances so leached out 163 is well illustrated in the salts of sea-water, which rep- resent the generalized result of countless ages of this leaching process." This statement is found in one of the most distinguished of his valuable publications — "Rela- tion of Soils to Climate." The error of taking sea water as an indication of the relative proportions of elements leached from soils is suggested by the great variation in the composition of the salts found in the different seas. This variation is set forth by the great number of analyses that have been made of the waters of all seas, certain typical ones being given in the following tables: COMPOSITION OF THE SALTS IN SEA WATERS. A. — OPEN SEAS. Atlantic . . Pacific . . . North Sea B. — SUB-CLOSED SEAS. Mediterranean Baltic Sea Black Sea 0. — CLOSED SEAS. Caspian Sea Dead Sea . . . Si O2 trace trace trace trace trace CaO Per cent. Mg O Na, O CI Per cent Per cent. 0.045 0.09.51.108 0.047 131jl. 026 0.0320.1581.020 0.0048 0.3001.068 0,0036 0.1610.589 0.01.30 O.O661O. 551 SO., Per cent. Per cent. 1.946 257 1.895 0.278 1.8160.259 2.1090.571 1.038 0.072 957 0.125 0.019 0.040 0.114 0.273 0.137 0.215p.417|0.088 1.762 0.024 The composition of the waters of those several seas may be allowed to suggest the different composition of the rocks and soils which form the great water-sheds discharging into those seas. The suggestion must be taken with reserve, however, and before discussing it further we shall consider another argument showing 164 why the compositiou of sea-water cannot represent the relative amounts of the elements removed from the land and carried into the ocean by water. In the first part of these investigations we gave the composition of Hawaiian lavas, and also, in another place, the composition of the great mass of Hawaiian soils. For the present puii:>ose we shall bring the analy- ses of the lavas and soils together. The analyses of the soils are absolute, and give the full composition of the mineral matter of the soils. Hawaiian. SiO, Per cent. Lavas 47.900 Soils I 29 843 Al, O., Per cent. 18.230 27.221 Fe=0, Per cent. CaO Per cent. 13.360 8 990 34.326 0.928 MgO Per cent. 6.050 1.407 Na^O Per cent. 2.200 1 641 K„0 Per cent. 1.500 0.853 This comparison shows that, in the jiassing over of the lavas into soils, there have been removed 18 tons, out of every 48 tons, of silica; 8 tons out of every tons, of lime; 4^ tons, out of every 6 tons, of magnesia; one-half of every 1^ tons of potash; and only one-fourth of every 2^ tons of sodium. We see then, that sodium, which constitutes only about 2 per cent, of the original lava, is removed in the least proportion of the elements speci- fied. This behaviour of sodium is in keeping Avith all our other observations, and with what we know of the sili- cates of the metal in the arts and manufactures, which are conspicuous by their insoluble nature; and it in no- wise conflicts with the further observation, that when sodium is once made soluble, it is removed from soil more rapidly than potash, for example, which is due to the different relations of the elements to the absorptive property of soil. 165 In considering the disintegration of Hawaiian lavas, and the elements that are removed, we have to bear in mind their difference of composition as compared with other rock-formations that are in constant course of de- cay, and whose elements are being carried elsewhere, and into the sea. We say "elsewhere," for the reason that, in the decomposition of original lavas, elements are separated out, which go to the making of deposits, in large mass, at lower altitudes, whose elements do not go direct to the sea, yet they are not accounted for in general soils. These results of disintegration were considered in the first part of this work. It is hardly a matter, however, that is difficult to ex- plain w^hy so much lime, magnesia, potash and silica, and so little soda, have been carried into the sea, and yet sodium compounds compose such a vast proportion of the salts in most ocean waters. Lime, and also mag- nesia with mineral acids, constitute the material with which those representatives of the animal kingdom liv- ing in the waters of the seas have built up their structures of bone and shell, and which bones and shells have gone towards the laying down of those massive formations of limestone which are found upon so grand a scale in the structure of the earth. The lime of those limestones was at one time bound up in the composition of rocks and lavas. On their disintegration. It was carried into the seas; and from the seas, it was gathered up by the myriad denizens of the ocean and stored away in the masses in Avhich it is now found. The millions of tons of coral reef which encompass these islands, and which our analy- ses show to contain over 96 j)er cent, of lime carbonate, and one or more per cent, carbonate of magnesia, form a most elaborate instance of what becomes of the lime. IGG Then tlie vegetable kiugdom flonrishiiig b(nieath the sur- face of the seas, — its orders draw upon the nitrogen, phos- phoric acid, and still more upon the silica, and also the potash, which have gone from the decaying rocks and soils into the sea. These indications of the behaviour of the potash have a very particular bearing upon the views hitherto set forth by agricultural chemists. The sum of these things, therefore, causes us to look upon the salts found in the waters of seas not as the collective mass of material which was carried out from the land, but rather as the residue of matter remaining after the animals and plants lirinfi and multlplyinf/ in the seas, hy their selective action, hare talrn out of the n-aters thi- jiiissiiiff elements, each animal and plant after ifx manner (Did needs. Having concluded that the composition of sea water does not furnish any means of estimating the relative proportions in which the elements composing rocks and soils are removed and carried into the sea, w^e undertook an examination of the actual iraters of discharge which are leaving the lands at this time. To ascertain the composi- tion of those waters at the present time will amply serve our purpose; yet it must be borne in mind that the com- position of the waters that are discharging to-day will not indicate the relative amounts of the elements that have been borne to the sea during previous periods of time. When original rocks and lavas commence disinte- gration, in the first stages of the process, the more solu- ble elements are removed first, and in great excess. As the decomposition proceeds, and little of the most solu- ble is left to be removed, the less soluble elements must come more into prominence in the composition of the discharging streams. Consequently not only the nature of the rocks, but also their age and state of decay, have a 167 bearing upon the composition of the material that is being removed from them by water. Before treating of the waters discharging from the Hawaiian Islands, we shall consider snch data as we possess bearing upon the character of the waters dis- charging from lands in older countries. Unfortunately we are not sure (although it is probable), that the data represent waters in normal discharge, or whether they may not be from streams at a time when the volume of discharge was either less or greater than normal, which diiferences can affect the composition of the water. Again, few analyses of water have been made from the standpoint of our present purpose. The reasons for water analysis generally, are hygienic, and not geologi- cal; and the examinations are seldom full. Judging ^'•'»ni data that are available, more attention has been given to water analysis in England than in other of the older countries. We therefore bring together in average such analyses as bear upon our purpose. This average embraces the composition of streams and springs in the southern counties of England — Middlesex, Kent, etc., and is as follows: Si O.. Per cent. English waters 0008 Fe, Al, O 2 Al2 ^6 Per cent. 0.0002 Ca O Per cent. Mg O Per cent. 0.0096 0004 Ka O Per cent. 0002 Na, O Per cent 0.0007 01 Per cent. 0.0009 Attention is first called by the analysis to the amount of silica in the discharging waters. This is very note- Avorthy, since in sea waters merely a trace is recorded. It is seen that the lime being carried to the sea is fourteen times greater than the soda, the great excess of lime be- 108 iiiji (hie to the huge formations of limestone found in Southern England, and through which the drainage has passed. Even the potash is one-third of the quantity of the soda in those waste waters, whilst in sea water, ac- cording to Regnault, there are only 2 parts of potash to 100 parts of soda. We shall not use more time in dis- cussing the features of composition of the discharge waters in other countries. The examples used are only given for their value in comparison, and to illustrate the truth that the composition of the mineral matter in waste waters is dependent upon the nature and age of the rock- formations over and through which the waters are flow- ing. The composition of the waters discharging from the four larger islands of the Hawaiian group will now be considered, and in geological relation to the main pur- pose of these investigations. As soon as it appeared to the writer that a knowledge of the composition of the mineral matter in the waters discharging from the four chief islands into the sea would be absolutely necessary to a solution of matters that form an integral part of our main investigation, a course was adopted by which waters were selected, and samples taken, that should represent the water-sheds of all the great mountains that are discharging their waste waters through the soils and lava formations of each district of the four islands — Hawaii, Maui, Oahu and Kauai. The areas of these islands are, in the order of the names given, respectively 4,210, 7G0, 000 and 590 square miles, or an aggregate of G,160 square miles, or 3,110,000 acres. These areas indicate that the project Avas not only feasible, but that the conditions, sucli as the existence of the aggregate area in sub-areas of each island, specially 169 contribute to make possible tlie sampling* of the bulk of waste waters with convenience and exactness. In the course of the repeated visits to the islands, and the slow, methodic tours of study and observation, through each district of every island, the writer had opportunities to note the location, number, and size of the rivers, streams, and many of the springs of discharge, from several of which the ciibic-second volume has been determined. By means of these opportunities the loca- tions were selected for taking samples, which were as follows : Hawaii. — Kegion of Kohala; from the Kohala water- shed discharging in the district of Halawa and Niulii. Eegion of Hamakua; representing the main watershed of Mauna Kea, discharging at Pauilo and Kukaiau. Region of Hilo; the Wailuku river, being the main discharge from the slopes of Mauna Loa. Maui. — The great watershed of East Maui, discharg- ing by way of the slopes of the greatest crater mountain on the earth — Haleakala — samples taken from the Paia and Spreekelsville streams. Also from the watershed of West Maui, the samples being taken at Wailuku. Oahf. — From the streams discharging at Honolulu and Pearl Harbor. Kauai. — From the watershed on the north side of the island in the district of Kilauea. From the region of Waialeale, and mountains above Mahaulapu, discharg- ing at Koloa, and from the Waimea river, which stream gathers up the discharges from countless springs and streams that descend from the mountain ranges of the large region of Waimea. Our analyses of individual waters indicate very small variations in their composition, excepting samples taken 170 from locatious that are affected bj the sea, in which not only the sodium increases, but the proportions of the other bases have become changed. The iiraud average, which sets forth the composition of the mineral matter found in the waters discharging by way of the watersheds, and from all the areas speci- fied, is given in the following data of analysis, under- neath A\ liicli we reproduce the com])osition of the English waters for tlie instruction afforded by the comparison: Waters. Si Oo Fe. Al„ Oe CaO Mg-O K„0 Na=0 Cl Per cent 0.0040 SO3 Per cent. 0.0011 P2O6 Hawaiian Per cent. 0023 0.0008 Per cent 0.0005 Per cent. 0.0013 Per cent 0.0013 Per cent. 0.0005 Per cent. 0.0033 Per cent. 0,0001 English . . . 0.0002 0.0096 0.0004 0.0002 0.0007 0.0009 Passing on from the great interest and instruction that are afforded by the comparison of the analyses, the nature of which we have discussed in previous pai*a- graplis, we come into the possession of data which fur- nish the most reliable means that we can use in estinmt- ing the relative amounts of the elements of plant food that are removed from the soil by the processes of Na- ture, and of bringing the results of the laboratory for com])arison, by the side of truths obtained from the field, which facts and conclusions found in the ways of Nature must be a standard in judgment. Before pro- ceeding Willi the comparison, we shall try to supplement the conclusions furnished by the study of the tenters of discharf/c with other data strictly furnislied by the field. Removal of the Elements of Plant Food by Cropping. — In speaking of the removal of plan food materials by ''cropping," we have very carefully io distinguish be- 171 tween the total amount removed by acts of cultivation, by rain falling on cultivated land, and by crops, all of which make up the total action of "cropping," and the small amounts that are taken away by the crops alone. We have data showing that of 7000 lbs. of lime removed from given lands per acre by cropping, not quite 1000 lbs. of that amount were carried off by the actual crops. Also that not more than one-lialf of the potash re- moved was actually taken away in the harvested growths. These data also show that any system of judging of the depletion, or of restoring, the fertility of soils, that is based upon a mere calculation of the amounts of the elements that are carried away from the land in crops is devoid of any approach to the actual facts in the matter. In the course of the past three years we have taken samples of soil from all districts on each of the four larger islands. As already explained, those samples were taken personally by the writer, or strictly in locali- ties and fields selected by him, and according to his in- structions. Those soils have undergone the common agricultural analysis, making altogether analyses of 94 type soils, and 1328 sub-samples. Of this number, 64 type soils, and 768 sub-samples, were examined in such a way as to throw special light upon the present relative compositions of "virgin" and "cropped" soils. As it is said in the report on soils for 1895, "In taking samples of cropped lands, in every possible case, a sample of vir- gin soil was taken, corresponding in all requisite con- ditions to that of the cropped soil, and a comparison is made of the two." The 64 type soils represent both up lands and lowlands. For our present purpose we ar(f: able to use and consider only the upland soils, since the 172 question before ns is the "aniouuts of materials that have been removed from the soils by cropping," and this question only applies to the uplands, from which the elements have been washed away by the rains, and in some cases carried to the lowlands, and not direct into the sea, which is shown by some of the lowland cropped soils being richer in given elements than the virgin soils of the same localities. For this purpose there are some 34 type soils, including 408 sub-samples of soil. These type soils are strictly representative of the uplands, and were taken from most districts on each island. We have analyses of about 140 more sub-samples of cropped up- land soils, which correspond in results with the others, but as there are not virgin samples to compare with these, they are excluded from the comparison. In the following table we bring together in average the results of the 34 types, and 408 sub-samples: UPLANDS. Elements. Virgin. Cropped. Loss. Lime Per cent. 415 324 0.248 Per cent 0.248 0.270 0.243 Per cent. 40 20 Potash .... Phosphoric Acid 16 60 2.02 The results of the analyses, which varied between more and less than the grand average, show that 40% of the lime, 16.6% of the potash, and 27o of the phosphoric acid have been removed from the cultivated ii]>l;nid soils by cropping. The amount of phosj^horic acid rein(»ve