o:sr THE 
 
 Pho?i"P'^4c Gtiaiic Mends of the ¥hi'Ak» 
 
 BY J, ]). HAGUE. 
 
H 
 
 l;iss_ 
 
 %h¥i 
 
 hook 
 
 1 .' . /\ 
 
 SMITIIS()X:.\X DKPOSIT 
 
ON THE 
 
 Phosphatic Guano Islands of the Pacific. 
 
 
 <vv-,.,^,,'Y\^;' 
 
[From the American Journal of Science and Arts, Vol. XXXIV, Sept. 1862.] 
 
 ON THE 
 
 PHOSPHATIC GUANO ISLANDS 
 
 PACIFIC OCEAN, 
 
 J. D. HAGUE. 
 
 During a few years past the attention of scientific men and of 
 agriculturists has been called to some varieties of Phosphatic 
 Guano found on several small islands of the tropical Pacific and 
 imported to this country and to Europe under the name of 
 " American Guano." 
 
 The principal ingredient of these guanos is the phosphate of 
 lime, with which is combined in the various sorts more or less 
 phosphate of magnesia, sulphate of lime, organic matter and 
 water. They generally contain traces of ammonia with a small 
 percentage of soluble salts, but these, which, without doubt, 
 formed an important part of the guano as it originally existed, 
 have now almost entirely disappeared in consequence of the va- 
 rious changes to which the deposits have been subjected, 
 
 * Much of the chemical investigation of which the results are given in this paper 
 I made in the Siieffiekl Laboratory of Yale College, the facilities of which were 
 kindly afforded me by my friends, Profs. Brush and Joiinson, to whom I am happy 
 to express my thanks for this favor, and for their valuable assistance in the prosecu- 
 tion of my work. Also to my brother, Mr, Arnold Hague, one of their students, my 
 acknowledgments are due for analytical aid. j. d. h. 
 
J. D. Hague on the Guano Islands of the Pacific Ocean. 3 
 
 The first samples of these guanos were taken from Jarvis' and 
 Baker's Islands in 1855 and sent to the United States for exam- 
 ination, the results of which led in 1858 to the occupation and 
 working of the deposits. The importance and value of these 
 having once become evident, the Pacific, within a few degrees 
 north and south of the equator, was carefully explored and 
 many other islands were visited, on a few of which beds of gu- 
 ano of some extent were discovered. 
 
 In the following paper I propose to describe some of these. 
 I shall have reference chiefly to Baker's, Rowland's and Jarvis' 
 Islands, on each of which I resided several months for the pur- 
 pose of studying the character and formation of their deposits. 
 I also spent some months in exploring this region of the Pacific 
 and visiting many other islands, having a small vessel employed 
 especially for that object. In this service, altogether, I was en- 
 gaged more than two years, from 1859 to 1861 inclusive, in the 
 employ of William H. Webb, Esq., of New York, by whose 
 courtesy I am permitted to publish these results. 
 
 These islands are all of coral formation. They are situated 
 near the equator and between the meridians of about 155° and 
 180° longitude west from Greenwich. They are without fresh 
 water and almost entirely destitute of vegetation, and are the 
 resort of countless thousands of birds whose accumulated ordure 
 and dead bodies have formed extensive deposits. 
 
 Baker's Island. — This island possesses the most important of 
 these deposits. It is situated in lat, 0° 13' north and long. 176° 
 22' west from Greenwich. Excepting Howland's Island, forty 
 miles distant, it is very remote from any other land. It presents 
 the usual features of an ordinary coral island. It is surrounded 
 by a fringing reef, which is from 200 to 400 feet wide and 
 slightly elevated above the sea level at low tide. It is about 
 one mile long and two-thirds of a mile wide, trending east and 
 west. The surface is nearly level, the highest point of which is 
 twenty-two feet above the level of the sea, showing some evi- 
 dences of elevation,* 
 
 * The accompanying engraving exhibits a section of the western (lee) beach ■which 
 was cut through for a railway. LL is the level of the reef of wliich the seaward 
 end P is the siiore platform or plateau covered at high tides by five and a half feet 
 
 of water. From the shore to the edge of the guano deposit G, is from ,S00 to 400 
 feet. The perpendicular height from LL to tlie summit of the sand beach, SS, ia 
 twenty-two feet, and the depth of the excavation opposite this highest point is ten 
 feet, tlie drawing being a little out of proportion. 
 
 The dotted line, «6, represents an old beach formation which the cut exposed. It 
 consists of large and small coral fragments and shells beneath which the sand lies in 
 compact strata. This formation was evidently once the surface of the island, and 
 
 Am. Jour. Sci.— Second Series, Vol. XXXIV, No. 101.— Sept., 1S63. 
 29 
 
4 J. D. Hague on the Guano Islands of the Pacific Ocean. 
 
 Above the crown of the beach there is a sandy ridge which 
 encircles the guano deposit. This marginal ridge is about one 
 hundred feet wide on the lee side of the island, and is there com- 
 posed of fine sand and small fragments of corals and shells mixed 
 with considerable guano; on the eastern or windward side it is 
 much wider and formed of coarser fragments of corals and shells 
 which, in their arrangement, present the appearance of successive 
 beach formations. This margin is partially covered with a rank 
 growth of long, coarse grass, portulacca, mesembryanthemum, 
 and a few other species of plants. 
 
 Encircled by this ridge lies the guano deposit occupying the 
 centre and the greater part of the island. The surface of this 
 deposit is nearly even, but the hard coral bottom which forms 
 its bed has a gradual slope from the borders towards the centre, 
 or, perhaps more properly, from northwest to southeast, giving 
 the guano a variable depth from six inches at the edges to sev- 
 eral feet at the deepest part. None of the grass that grows 
 abundantly on the margin is found on the guano, but there are 
 one or two species of portulacca occurring in certain parts, (par- 
 ticularly where the guano is shallowest and driest), and to this 
 is owing the presence of the fine roots and fibres in some of the 
 guano. 
 
 The entire deposit presents considerable uniformity in charac- 
 ter. Excepting some isolated spots of little extent there is no 
 outer crust, and the guano of the surface differs but little, if any, 
 from that below. There is, however, some variety in the ap- 
 pearance of the guanos of the deep and shallow parts of the de- 
 posit. On the northern side it is from six to twelve inches deep ; 
 is generally quite dry, and is a dark brown pulverulent substance 
 of rather coarse grain or texture, containing many thread-like 
 roots and fibres and whitish particles, among which Prof Liebig 
 observed scattered crystals of the phosphate of magnesia and 
 ammonia.* It is closely though not hard packed, and is readily 
 
 may be traced from a to 6, -where the guano rests upon it. Above it lies a sandy 
 ridge, SS, a comparatively new beach accumulation ralher indistinctly stratified. 
 The highest point of ab is fifteen feet above LL, which altitude, in accordance with 
 the commonly accepted theory that the sea-made coral land does does not exceed 
 ten feet in height, would, of itself, be an evidence of elevation and, consequently, to 
 account for the present height of twenty-two feet, it would be necessary to suppose 
 a subsequent subsidence in order to allow SS to accumulate, and finally another 
 elevati()n of the whole to its present position. It must be observed, however, tliat 
 the sandy ridge, SS.only prevails at this altitude on the southwestern shore, and 
 probably violent westerly gales and heavy seas have had much to do with its forma- 
 tion. My own observations favor the opinion that the sea-made coral land may 
 reach a greater altitude than ten or twelve feet. During the prevalence of high 
 6urf at Jarvis Island I have known seas to wash up the beach with body and force 
 eufficient to carry away plank and ppars that were lying on the crown of the beach 
 eighteen feet above the level of the reef. 
 
 * Liebig's Keport on Baker and Jarvis Guanos, Aug. Ith, 1860. 
 
/. D. Hague on the Guano Islands of the Pacific Ocean. 5 
 
 removed bj shovels without the aid of picks. Iq this part of 
 the deposit the portulacca flourishes most. 
 
 Tiie guano oti the southern side is of reddish color, of finer 
 texture, much damper, and of less specific gravity than that just 
 described. There is much less vegetation in this part of the 
 deposit, and the guano here contains scarce any roots or fibres. 
 
 Chemicall}' these varieties do not differ very much. Usually 
 the darker sort contains less water and more organic (vegetable) 
 matter, from which it probably derives its color. 
 
 Analyses of these two sorts are given beyond. 
 
 Much light may be thrown on the formation of these deposits 
 by the analysis, (I) which follows, showing the composition of 
 recently deposited guano. The sample itself does not represent 
 any considerable part of the existing deposit, but was taken 
 from a locality where large numbers of birds are still accustomed 
 to congregate. It is the dung of the Pelican us Aquilus, com- 
 monly called the Frigate Bird, which of all the birds frequenting 
 the island is the only one whose recent evacuations are of such 
 a consistency that they may conveniently be collected. They 
 contain a large proportion of solid matter, while the evacuations 
 of nearly all the other birds are very thin and watery. It is 
 found in their favorite roosting places, and shows the character 
 of guano before it has long been subjected to the influence of the 
 weather. It is a light and dry substance, consisting of friable 
 grains or fine powder, of a brown color, smelling strongly of 
 ammonia. Of the three following analyses No. I is this freshly 
 deposited guano; No. II is of the light colored guano from the 
 deeper part of the deposit, and No. Ill of the dark guano from 
 
 the shallow part. 
 
 I. II. III. 
 
 Moisture expelled at 21 2<' Fahr., 10-40 2-92 182 
 
 Loss by ignition, 36-83 8-32 850 
 
 Iiisol. ia HCl, (uiiconsumed by ignition), -78 ... ... 
 
 Lime 22-41 42-74 42'34 
 
 Magnesia, 1-46 2 54 2 75 
 
 Sulphuric acid 2 36 1-30 124 
 
 Phosphoric acid 21-27 39-70 4014 
 
 Carbonic acid, chlorine and alkalies, undet.,. . 4-44 248 3-21 
 
 lOU-00 10000 10000 
 
 Sol. in water remaining after ignition, 3 63 
 
 No. I contained 3'82 per cent of actual ammonia and all con- 
 tain traces of iron. I also obtained in sample I. a strong reaction 
 for uric acid. 
 
 This sample (No. I) resembles Peruvian guano in many re- 
 spects, and leads to the conclusion that the difference between 
 that and the American guano is mainly owing to circumstances 
 of climate. 
 
 In some parts of the deeper deposit a light scale or crust has 
 formed over the surface, which is generally very thin though 
 occasionallj'' hard pieces are found varying from half an inch to 
 
6 J. B. Hague on the Guano Islands of the Pacific Ocean. 
 
 an inch in thickness. The thin scale is met with particularly 
 where there is, or has been, any moisture, and, after showers, 
 where pools of water have been standing for some time, such a 
 crust appears on drying. There seems to have been a similar 
 process in the formation of the thicker crust, for it is found only 
 occasionally in places of which the dampness and general appear- 
 ance indicate that water may have assisted at its formation. 
 
 The thinner pieces are found not only on the surface, but in 
 certain localities form strata at various depths, usually about 
 an inch apart, with intermediate layers of guano. These strata 
 seem to have been formed at intervals during the accumulation of 
 the guano deposit each one at sometime having itself formed the 
 surface and now marking a period in its age. 
 
 Each of the localities where these strata occur, although on 
 opposite sides of the deposit are at the edges and immediately 
 adjoining the marginal ridge already described and from their 
 proximity to the shore it seems possible that these may have been 
 subjected to occasional floods by high seas washing over the 
 crown of the beach. 
 
 The following is an analysis of a thick and hard piece of crust 
 found on the surface : — 
 
 Loss by ignition (water and little organic matter) 11 "7500 
 
 Lime 4093 
 
 Magnesia "74 
 
 Phosphoric acid 40"47 
 
 Sulphuric acid 6"66 
 
 Loss and undetermined "45 
 
 10000 
 
 The small amount of magnesia and the excess of sulphuric 
 acid are points worthy of notice. 
 
 This crust is formed on Baker's Island only to a limited extent, 
 but its existence there and character are interesting when com- 
 pared with the Jarvis Island deposits, the better part of which is 
 all crust and in which, as Johnson and Leibig have observed, 
 much of the phosphoric acid is combined as the neutral phos- 
 phate of lime. The same is true of this crust of Baker's Island. 
 
 Before referring to the climate, birds etc. of this island, I will 
 first give some description of Rowland's and Jarvis' Islands. 
 
 HoiolancVs Island. — About forty miles in a north northwest 
 direction from Baker's, is situated Rowland's Island in lat. 0° 51' 
 north and 176° 32' west from Greenwich. It is about a mile and 
 a half long by a half mile wide, containing, above the crown of 
 the beach, an area of some 400 acres. The highest point is seven- 
 teen feet above the reef and ten or twelve feet above the level 
 of the high tide. It trends N.N.W. and S.S.E. The general 
 features of the island resemble those of Baker's. Its surface, at 
 least on the western side, is somewhat depressed and much of it 
 is covered by a growth of purslane, grass and other vegetation 
 
/. D. Hague on the Guano Islands of the Pacific Ocean. 7 
 
 like that on Baker's Island, but considerably more abundant. 
 Near the centre of the island there are one or two thickets of leaf- 
 less trees or brushwood, standing eight or ten feet high and oc- 
 cupying an area of several acres. The tops of these trees, in 
 which the birds roost, are apparently quite dead but the lower 
 parts near the roots, show signs of life after every rain. The 
 windward side of the island is formed by a succession of ridges 
 com[)osed of coral debris with some sand and shells, running 
 parallel to the eastern beach, each one of which may, at earlier 
 stages of the island's growth, have successively formed the weath- 
 er shore. Occasionally among these ridges a sandy bed is met 
 with in which some little guano is mixed. On the lee side there 
 is also a sandy margin of considerable width. Bits of pumice 
 and pieces of driftwood are scattered all over the island's surface. 
 The main deposit of guano occupies the middle part of the isl- 
 and and stretches, with some interruptions of intervening sand, 
 nearly from the north to the south end. Its surface is even and 
 in many places covered by a thick growth of purslane whose 
 thread-like roots abound in the ^uano where it grows. The de- 
 posit rests on a hard coral bottom and varies in depth from six 
 inches to four feet. The fact, already observed at Baker's, that 
 vegetation flourishes most where the guano is shallow is also 
 quite apparent here and the consequent characteristic difference 
 between the guano of the deep and shallow parts is distinctly 
 mnrked. The first variety, from the deeper part, is a fine pul- 
 verulent substance of reddish brown color, usually a little damp 
 in its native bed and almost quite free from roots or fibers. The 
 latter is of rather coarser texture, quite black and containing 
 many delicate roots and fibers and much vegetable matter. The 
 following analyses exhibit their comparative quality. No. 1 is 
 of the deep part, No. 2 of the shallow part of the deposit. 
 
 No. 1. No. 2. 
 
 Moisture at 212° Fahr 1-83 4-12 
 
 Loss by ignition 8 65 22"63 
 
 Insol. in HCl (unconsumed organic) matter 1-95 2'00 
 
 Lime 42- S6"90 
 
 Magnesia 265 ]-24 
 
 Sulpliuric acid 1-33 "68 
 
 Phosphoric acid 39-65 30-80 
 
 Carb. acid, chlorine and alkalies undeterm'd, 1 -94 1 -67 
 
 lUO-00 100-00 
 
 It will be seen that the main difference in these samples is in 
 the volatile matters present. Discarding the water and the or- 
 ganic matter, comparative analyses of the ash would vary but little. 
 
 Some interesting pseudomorphs occur buried in the guano of 
 this island. Coral fragments of various species were found that 
 Lad longbeen covered up under the deposit and in some of which 
 the carbonic acid had been almost entirely replaced by phos- 
 phoric acid. In such I have found seventy per cent phosphate 
 
8 J. D. Hague on the Guano Islands of the Pacific Ocean. 
 
 of lime. In many others the change was only partial and, on 
 breaking some of these, in the centre was usually found a nucleus 
 or core of coral still retaining its original hardness and composi- 
 tion, while the external parts had been changed from carbonate to 
 phosphate which, though soft and friable, still preserved the 
 structure and appearance of the coral, 
 
 Jarvis' Island. — Jarvis' Island is situated in lat. 0° 22' south 
 and long. 159° 58' west from Greenwich. It is nearly two miles 
 long by one mile wide, trending east and west, and containing 
 about 1000 acres. Like Baker's and Howland's it has the gen- 
 eral features of a coral island, but it differs from them essentially 
 in the fact that it once contained a lagoon which has gradually 
 been filled up with sand and detritus, while the whole island has 
 undergone some elevation. It therefore presents a basin-like 
 form, the surface being depressed from the outer edge towards 
 the centre. It is encircled by a fringing reef, or shore platform, 
 about 800 feet wide; from this a gradually sloping beach re- 
 cedes, the crown of which is from eighteen to twenty-eight feet 
 high, forming a ridge or border, of varying width, which sur- 
 rounds the island like a wall, from the inshore edge of which 
 the surface of the island is gently depressed. 
 
 Within this depression there are other ridges, parallel to the 
 outer one, and old beach lines and water marks, the remaining 
 traces of the waters of the lagoon, marking its gradual decrease 
 and final disappearance. 
 
 This flat depressed surface in the centre of the island is about 
 seven or eight feet above the level of the sea. It bears but little 
 vegetation, consisting of long, coarse grass, mesembryanthemum, 
 and portulacca, and that is near the outer edges of the island 
 where the surface is formed of coral sand mixed with more or 
 less guano. In the central and lower parts the surface is com- 
 posed of the sulphate of lime, and it is on this foundation that 
 the principal deposit of guano rests. This feature of Jarvis' 
 Island is an important one to consider in studying the difference 
 between the guano found on it and that on Baker's Island, for it 
 readily explains the presence, in much of the Jarvis Guano, of 
 the great excess of sulphate of lime, remarked by all who have 
 investigated it, while the unequal mechanical mixture of the 
 guano with the underlying sulphate accounts for the lack of uni- 
 formity in different samples. 
 
 In examining the foundation of the guano deposit on Baker's 
 or Rowland's Islands, by sinking a shaft vertically, the hard 
 conglomerate reef rock is found directly underlying the guano, 
 Eesting on this foundation the guano has undergone only such 
 changes as the climate has produced. On Jarvis' Island, how- 
 ever, after sinking through the guano, one first meets wiih a 
 stratum of sulphate of lime (sometimes compact and crystalline, 
 sometimes soft and amorphous) frequently two feet thick, beneath 
 
J. D. Hague on the Guano Islands of the Pacific Ocean. 9 
 
 "which are successive strata of coral sand and shells deposited one 
 above the other in the gradual process by which the lagoon was 
 filled up * 
 
 Of the origin of this sulphate of lime there can hardly be any 
 doubt. As the lagoon was nearly filled up, while, by the grad- 
 ual elevation of the island, the communication between the outer 
 ocean and the inner lake was constantly becoming less easy, 
 large quantities of sea water must have been evaporated in the 
 basin. By this means deposits would be formed containing 
 common salt, gj^psum and other salts found in the waters of the 
 ocean. From these the more soluble parts would gradually be 
 washed out again by the occasional rains, leaving the less solu- 
 ble sulphate of lime as we find it here. 
 
 Some additional light is thrown on this matter by the different 
 parts of the surface, which, though nearly flat, shows some slight 
 variety of level. The higher parts, particularly around the outer 
 edges, are composed chiefly of coral sand, either mixed with or 
 underlying guano. Nearer the centre is a large tract, rather 
 more depressed, forming a shallow basin in which the bulk of 
 the sea water must have been evaporated, and whose surface 
 (now partly covered with guano) is a bed of sulphate of lime, 
 while, further, there is a still lower point, the least elevated of 
 the whole, where the lagoon waters were, without doubt, most 
 recently concentrated. This latter locality is a crescent shaped 
 bed, about 600 feet long by 200 or 800 feet wide, having a sur- 
 face very slightly depressed from the outer edge towards the 
 middle. Around the borders are incrustations of crystallized 
 gypsum and common salt, ripple marks and similar evidences of 
 the gradually disappearing lake. The whole is composed of a 
 crystalline deposit of sulphate of lime, which, around the borders, 
 as already observed, is mixed with some common salt, while near 
 the centre, where rain water sometimes collects after a heavy 
 shower, the salt is almost entirely washed out, leaving the gyp- 
 sum by itself It is closely, but not hard, packed, and is still 
 very wet. By digging 18 or 24 inches down, salt water may 
 generally be found. 
 
 These facts help us to understand the varying conditions in 
 which we now find the guano beds, since the most important 
 part, and that from which the importations have thus far come, 
 rests on a bed of sulphate of lime, of an earlier but similar origin 
 to that just described above: a part rests on a coral formation, 
 while still anothei part, covering a large tract, has been by the 
 action of water mixed with coral mud. 
 
 The first named deposit, lying on the sulphate of lime bed, 
 has a peculiar character. It is covered by, or consists of, a hard 
 
 * These horizontal strata were penetrated to a depth of about twenty feet. 
 They were composed chiefly of fine and coarse sand with an occasional stratum of 
 coral fragments and shells. 
 
10 J. D. Hague on the Guano Islands of the Pacific Ocean. 
 
 crust that is from one-fourth of an inch to an inch and a half in 
 thickness, beneath which lies a stratum of guano varying in 
 depth from one inch to a foot. In many places where the guano 
 was originally shallow the whole is taken up and formed into 
 the hard crust which then lies immediately on the sulphate. 
 This crust, when pure, is snow-white, with an appearance some- 
 what resembling porcelain, but is usually colored more or less 
 by organic matter. Generally it is very hard, and strongly co- 
 hesive, though sometimes friable, and it lies unevenly on the 
 surface in rough fragments that are warped and curved by the 
 heat of the sun. It consists chiefly of phosphoric acid and lime, 
 but, owing to the variable amount of sulphate of lime with which 
 it is mechanically mixed, there is a lack of uniformity in differ- 
 ent samples. Hence the percentage of phosphoric acid varies 
 from over 50 per cent to less than 80 per cent. 
 
 The phosphoric acid and lime, moreover, are not combined in 
 constant proportions, some existing as bone phosphate, the 
 greater part, doubtless, in most specimens, as the neutral phos- 
 phate, and, possibly, a part as the superphosphate. 
 
 The following is an analysis of a piece of pure crust. The 
 sample, in question, was a snow-white fragment, containing 
 scarcely any organic matter. 
 
 Moisture at 212° Fahr -12 
 
 Loss by ignition, (combined water with little organic matter), 9-62 
 
 Lime, 3832 
 
 Sulphuric acid, 1'63 
 
 Phosphoric acid 50 04 
 
 Undetermined and loss '27 
 
 100-00 
 
 This presents a somewhat remarkable character. It appears 
 to be a nearly pure di-phosphate of lime. After allowing to the 
 sulphuric acid the requisite amount of lime, there remains 
 enough of the latter to form ninety per cent of the salt 2CaO, 
 HO, PO5 leaving an excess of about three per cent of phos- 
 phoric acid, which would suggest the possibility that a part of the 
 phosphoric acid and lime may be combined as CaO, 2H0, PO5. 
 
 So small an amount of sulphuric acid is also noticeable in a 
 specimen of Jarvis guano which usually contams a large per- 
 centage of that acid, but in this case it is owing to the purity of 
 the crust and the absence of mechanically mixed sulphate of lime. 
 
 Samples of Jarvis guano have been examined by many chem- 
 ists, but their results are not always uniform, because, as I have 
 already explained, their samples were mixtures of this crust and 
 the underlying guano or gypsum. A number of analyses, made 
 for commercial purposes by Prof Johnson of New Haven, I find 
 published in a guano pamphlet, issued by Mr. Webb as a trade 
 circular. Prof Liebig has also published a very complete analy- 
 
/. D. Hague on the Guano Islands of the Pacific Ocean, 1 1 
 
 sis of Jarvis guano in liis "Keport on the Guanos of Baker's 
 and Jarvis' Islands, Aug. 7th, 1860." 
 
 The following presents some of the results obtained by these 
 two chemists : 
 
 Liebig. Johnson. 
 
 Average of four samples. 
 
 Lime 34-839 34-79 
 
 Phosphoric acid, 17-601 18-48 
 
 Sulphuric acid, 27 021 20-75 
 
 In Johnson's samples nearly the whole of the phosphoric acid 
 is combined with the lime as 2CaO, HO, PO^, while Liebig finds 
 for the above, 
 
 3CaO, PO5,.,. , 17-397 percent. 
 
 2CaO, HO, PO5, 16026 " 
 
 The formation of the neutral phosphate in this guano I think 
 may be considered as a result of the action of sea water to which 
 this part of the deposit has been subjected. It will be remem- 
 bered that in describing the Baker's Island deposit I gave an 
 analysis of a piece of crust found there, in which the phosphoric 
 acid was likewise partly combined as the neutral salt. In that 
 crust was also noticed a much larger percentage of sulphuric 
 acid than is found in the guano from which it was formed ; and, 
 further, it was observed that on Baker's Island this crust only 
 occurs in places of which the appearance and position indicate 
 that water (probably from high seas washing over the crown of 
 the beach) assisted at its formation. It seems to me probable, 
 under these circumstances, that sulphates resulting from the 
 evaporation of the sea water have been decomposed, and that 
 the sulphuric acid has united with the lime of the bone phos- 
 phate, causing the formation of the di-phosphate of lime. 
 
 That this process may have been carried on to a much greater 
 extent at Jarvis' Island, where much of the deposit has evidently 
 long been acted upon by sea water, seems to me beyond a doubt. 
 
 A singular feature is presented by this crust in the formation 
 of so-called 'hummocks,' an idea of which may be better obtained 
 from the accompanying cuts than from words. These 'hummocks^ 
 
 
 vary in diameter from one to ten inches and in height from half 
 an inch to six or seven inches. The exterior is composed of the 
 
 hard, phosphatic crust, while within each one, without exception^ 
 there is a central mass of soft, amorphous and nearly pure hy- 
 drated sulphate of lime. When one of these is cut through ve.v- 
 Am. Jour. Sci.— Second Sebies, Vol. XXXIV, No. 101.— Sept., 1S63. 
 30 
 
12 J. D. Hague on the Guano Islands of the Pacific Ocean. 
 
 ticallj the section shows a series of concentric layers above and 
 around this central mass. The exterior is almost pure phosphate, 
 and, proceeding from the outside towards tlie centre, each suc- 
 cessive layer has less phosphate and more sulphate until the 
 central mass is reached, which is almost pure sulphate. It is 
 worthy of note that this hj^drated sulphate of lime, which inva- 
 riably fills the centre of a "hummock," is amorphous and ex- 
 ceedingly fine and soft, even when the underlying gypsum is 
 crystalline. These hummocks are scattered over certain parts of 
 the deposits and occur in close proximity to each other. In 
 these places the deposit is invariably damp, and, usually, be- 
 neath each one may be found, mixed with the underlying sul- 
 phate, a black, earthy and damp substance containing much 
 phosphate and some carbonate of lime. This black substance 
 was, probably, coral mud, in which, as in the coral pseudomorphs 
 of Howland's, the carbonic acid has been expelled and replaced 
 by phosphoric acid, and this affords the only explanation that I 
 can offer for this remarkable formation, namely, that in the 
 chemical interchange that must have taken place between the 
 soluble salts washed down from the guano on the surface, the 
 sulphate of lime and the coral mud, there may have been an ex- 
 cess of carbonic acid liberated from the latter and replaced by 
 phosphoric acid. The surface guano was probably wet and in a 
 plastic state like thick mud, and the ascending carbonic acid, 
 finding no other means of escape, and exerting an upward 
 force, produced these hummocks, which have since become dry 
 and hard. 
 
 In those parts of the crusted deposit where there are no " hum- 
 mocks" the surface is usually a little higher and the deposit be- 
 low drier than where the hummocks occur, and this would fur- 
 nish a reason for their absence, since the hummocks could hardly 
 be formed, as above explained, if the surface, for want of moist- 
 ure, were not sufficiently plastic and yielding. 
 
 Thus this guano has not only been deprived of its ammoniacal 
 salts, uric acid, etc., as have the deposits of Baker's and How- 
 land's, but by its immediate contact with the gypsum has under- 
 gone further chemical and physical clianges. Probably, too, the 
 direct action of sea water has effected much by brmging together 
 and mixing the guano with the bed on which it lay, and, by oc- 
 casional inundations, exposing the whole alternately to the ac- 
 tion of water and to the intense heat of the sun.* Thus it has 
 been baked into a thick and hard crust whose chemical compo- 
 sition differs materially from the guano in its usual form. 
 
 * McKean's and Phcenix Islands, described below, are likewise old lasfoons not 
 yet elevated so hi^li as Jarvis's. Their basins are sometimes flooded at high tides 
 by several inclies of water. Thus we may suppose that Jar vis, in an earlier stage 
 of the process of elevation, was subjected to occasional floods, keeping in mind the 
 fact before mentioned, that by digging now in the lower parts of the island salt wa- 
 ter may be found at no great distance below the surface. 
 
J". D. Hague on the Guano Islands of the Pacific Ocean. 13 
 
 I have said that there was beneath the crust a stratum of gu- 
 ano of variable depth. Frequently it is wanting altogether, the 
 whole being taken up in the crust and lying in immediate con- 
 tact with the bed of gypsum. Where there is such a layer of 
 guano it is variable in composition, being mixed with more or 
 less sulphate of lime. 
 
 It generally contains from sixty to seventy per cent phosphate 
 of lime. 
 
 I come now to speak of that part of the Jar vis deposit which 
 rests on a coral foundation. This is of limited extent, but is of 
 great interest because of its similarity to the Baker guano. It 
 is about two feet deep ; is a dry powder of dark brown color, of 
 rather lighter shade than the Baker guano, owing to the pres- 
 ence of less vegetable matter. It contains very little coral sand 
 mixed with it. The following is an analysis: 
 
 Moisture at 212° Falir., 5-02 
 
 Loss by ignition, 8'45 
 
 Lime, 42 17 
 
 Mas^nesia , 102 
 
 Sulphuric iicid 3'()6 
 
 Phosphoric acid, 34-01 
 
 Carbonic acid, '81 
 
 Iiisol. residui*, (organic matter unconsumed by ignition),. ..... •%{} 
 
 Chlorine, alkalies, iron, etc., 4 "86 
 
 100 00 
 
 It is important to observe that while the greater part of the 
 Jarvis guano, as already described, differs materially from the 
 Baker, this portion of the Jarvis deposit has almost the same 
 chemical and physical characteristics as the Baker or Plowland 
 guano. Resting like that on a coral foundation, it has been ex- 
 posed only to like influences, while the Jarvis crusted deposit, 
 above described, owes its peculiar character to its contact with 
 the gypsum on which it lies and to the action of the sea water. 
 
 This gypsum or sulphate of lime is usually soft and amor- 
 phous, sometimes crystalline, and, at a depth of eighteen inches 
 or two feet, occurs in hard, compact, crystalline beds. It is of a 
 light snuff color, and where it underlies guano, is mixed with 
 considerable phosphate of liine, which has been washed down 
 from the surface. Similar deposits of sulpliate of lime occur on 
 many other elevated lagoon-islands of the Pacific, someof Avhich 
 I shall allude to below. I have also seen gypsum, of similar 
 character and appearance, which occurs in "pockets" or small 
 depressions in the now elevated portions of the coral reef at 
 Oahu, Sandwich Islands, and doubtless due to the same source, 
 the evaporation of sea water. 
 
 Unfortunately for the commercial interests of the Jarvis guano, 
 the earlier cargoes (the first one or two) that were brought thence 
 were selected without the aid of chemical anal3'sis, and those in 
 charge mistaking the gypsum for guano, sent home cargoes, the 
 
14 /. D. Hague on the Guano Islands of the Pacific Ocean. 
 
 greater part of which was far from being worth the expense of 
 transportation. The repetition of this error was promptly 
 guarded against by sending a chemist to the island, but it re- 
 quired a longer time for the reputation of the article in the mar- 
 ket to recover from the ill effects of such a mistake. 
 
 Climate. — The climate of these three islands is similar and 
 very equable. The trade winds are almost constant, and blow 
 in the summer from east by south to southeast, and, in the win- 
 ter, from east by north to northeast. From October to February, 
 inclusive, on Baker's, I did not observe a point of southing in 
 the wind, while during the summer months there are long periods 
 during which the wind is invariably from south of east. Calms 
 are rare, especially those of long duration. Westerly winds have 
 seldom been observed, except, occasionally, as light puff's on quiet, 
 calm days. On one or two occasions only, in the winter, at Ba- 
 ker's, have any westerly winds of much force been recorded. 
 
 The sky is clear and cloudless. The temperature is exceed- 
 ingly even, ranging from 76° at sunrise to 88° Fahrenheit at the 
 hottest part of the day in the shade. In the sun at noon it 
 stands between 95° and 100°. 
 
 Eain falls in light showers not infrequently. Heavy showers 
 are rare and rainy days are unknown in my experience there. 
 During four winter months at Baker's Island, from October 1, 
 1859, to February 15, 1860, rain fell twenty-three times, gener- 
 ally occurring in light showers or squalls, at intervals of a week 
 or thereabouts, and a general coincidence between the times of 
 occurrence of these showers and the changes of the moon from 
 phase to phase has been observed, but this regularity is not so 
 great, neither at this or other seasons, but that weeks have 
 passed without a drop of rain. 
 
 During these four months the least of these showers, measured 
 by conical rain gauge, amounted to y/^ o of ^^ ^"^h on a level, 
 and the greatest on December 19, 1859, was yVA of one inch. 
 From December 14, 1859, to December 20, 1859, inclusive, there 
 fell -Yo of one inch. The total amount of the four months' rain 
 ■was 1-840 inches, of which yVo f^ll in December. 
 
 Although the amount of rain falling in the summer months is 
 much less than that which falls in winter, there are, nevertheless, 
 days in summer on which showers have fallen as heavy as any 
 in the year, 
 
 Eain falls most frequently in the night and just before day- 
 break ; sometimes by day, especially if the sky has long been 
 overcast, a rain cloud passes over the island, but I have often 
 observed the remarkable phenomenon of a rain squall approach- 
 ing the island, and just before reaching it, separating into two 
 parts, one of which passed by on the north, the other on the 
 south side, the cloud having been cleft by the column of heated 
 air rising from the white coral sands. 
 
/. D. Hague on the Guano Islands of the Pacific Ocean. 15 
 
 The position of these islands near the equator and their re- 
 moteness from any high land make them favorable places for 
 studj-ing the meteorology of this region. The equatorial cur- 
 rent is a matter of great interest. It has a general direction of 
 west southwest, and runs with a great velocity, sometimes ex- 
 ceeding two knots per hour, and, at times, suddenly changing 
 and running quite as rapidly to the eastward. 
 
 Daring the winter months there are days when the swell is 
 -very heavy, and the surf breaks violently on the reefs, but in 
 summer there is little or no surf, and especially on the lee side 
 of the island, the water is very smooth. These periods in the 
 winter occur usually at intervals of a few days and prevail dur- 
 ing two or three and sometimes more days. In this connection 
 I may allude to the shifting sands at Baker's, which, as I ob- 
 served there, change their place twice in the year. The western 
 shore of the island trends nearly northeast and southwest; the 
 southern shore east by north. At their junction there is a spit 
 of sand extending out towards the southwest. During the sum- 
 mer the ocean swell, like the wind, comes from the southeast, to 
 the force of which the south side of the island is exposed, while 
 the western side is protected. In consequence the sands of the 
 beach that have been accumulating during the summer on the 
 south side are all washed around the southwest point, and are 
 heaped upon the western side, forming a plateau along the beach 
 two or three hundred feet wide, nearly covering the shore platform, 
 and eight or ten feet deep. With October and November comes 
 the wdnter swell from northeast, which sweeps along the western 
 shore and from the force of which the south side is in its turn pro- 
 tected. Then the sand begins to travel from the western to the 
 southern side, and after a month or two nothing remains of the 
 great sanS plateau but a narrow strip, while on the south side 
 the beach has been extended 200 or 800 feet. This lasts until 
 February or March when the operation is repeated. 
 
 Birds, etc. — From fifteen to twenty varieties of birds may be 
 distinguished among those frequenting the island of which the 
 principal are Gannets and Boobies, Frigate Birds, Tropic Birds, 
 Tern, Noddies, Petrels, and some game birds as the Curlew, 
 Snipe and Plover. Of terns there are several varieties. The 
 most numerously represented is what I believe to be the Sterna 
 Hirundo. These frequent the island twice in the year for the 
 purpose of breeding. They rest on the ground, making no nests 
 but selecting tufts of grass, where such may be found, under 
 which to lay their eggs. I have seen acres of ground thus 
 thickly covered by these birds, whose numbers might be told by 
 millions. Between the breeding seasons they diminish consid- 
 erably in numbers, though they never entirely desert the island. 
 They are expert fishers and venture far out to sea in quest of 
 prey. The Noddies (Sterna stolida) are also very numerous. 
 
16 J. D. Hague on the Guano Islands of the Pacific Ocean. ' 
 
 They are black birds, somewhat larger than pigeons, with mucli 
 longer wings. They are very simple and stupid. They burrow- 
 holes in the guano in which they live and raise their young, 
 generally inhabiting that part of the deposit which is shallowest 
 and driest. Their numbers seem to be about the same through- 
 out the year. The Gannet and Booby, two closely allied species, 
 (of the genus Sula), are represented by two or three varieties. 
 They are large birds and great devourers of fish which they take 
 very expertly, not only catching those that leap out of water 
 but diving beneath the surface for them. They are very awk- 
 vpard and unwieldy on land, and may be easily overtaken and 
 captured if indeed they attempt to escape at all on the approach 
 of man. They rest on the trees wherever there is opportunity, 
 but on these islands they collect in great groups on the ground 
 •where they lay their eggs and raise their young. One variety, 
 not very numerous, has the habit of building up a pile of twigs 
 and sticks, twenty or thirty inches in height, particularly on 
 Howland's where more material of that sort is at hand, on which 
 they make their nest. When frightened these birds disgorge 
 the contents of their stomachs, the capacity of which is some- 
 times very astonishing. They are gross feeders, and I have 
 often seen one disgorge three or four large flying fish fifteen or 
 eighteen inches in length. 
 
 The Frigate Bird (Tachy petes Aquilus) I have already al- 
 luded to. It is a large rapacious bird, the tyrant of the feath- 
 ered community. It lives almost entirely by piracy, forcing 
 other birds to contribute to its support. These frigate birds 
 Lover over the island constantly, lying in wait for fishing birds 
 returning from sea to whom they give chase, and the pursued 
 bird only escapes by disgorging its prey, which the pursuer very 
 adroitly catches in the air. They also prey upon flying fish and 
 others that leap from sea to sea, but never dive for fish and 
 rarely even approach tlie water. 
 
 The above are the kinds of birds most numerously represented 
 and to which we owe the existing deposits. When the islands 
 were first occupied they were very numerous but have since 
 been perceptibly decreasing. 
 
 Besides these are the Tropic Birds which are found in con- 
 siderable numbers on Howland's Island, but seldom on Jarvis' 
 or Baker's. They prefer the former because there are large 
 blocks or fragments of beach rock, scattered over the island's 
 surface, under which they burrow out nests for themselves. A 
 service is sometimes required of this bird which may, perhaps, 
 be worthy of notice. A setting bird was taken from her nest 
 and carried to sea by a vessel just leaving the island. On the 
 second day, at sea, a rag, on which was written a message, was 
 attached to the bird's feet, who returned to the nest, bringing 
 with it the intelligence from the departed vessel. This experi- 
 
/. D. Hague on the Guano Islands in the Pacific Ocean. 17 
 
 merit succeeded so well that, subsequently, tliese birds were 
 carried from Howland's to Baker's Island, (forty miles distant), 
 and, on being liberated there, one after the other, as occasion de- 
 manded, brought back messages, proving themselves useful in 
 the absence of other means of communication. 
 
 There are several varieties of tern, those described above, 
 however, being the only kinds that are found in very considera- 
 ble numbers. The game birds, snipe, plover and curlew, frequent 
 the islands in the fall and winter, but I never found any evi- 
 dence of their breeding there. They do not leave the island in 
 quest of prey but may be seen at low tide picking up their 
 food on the reef which is then almost dry. 
 
 Some of the social liabits of these birds are worthy of remark. 
 The gannets and boobies usually crowd together in a very ex- 
 clusive manner; the frigate birds likewise keep themselves dis- 
 tinct from other kinds; the tern appropriate to themselves a 
 certain portion of the island; each family collects in its accus- 
 tomed roosting place but all in peace and harmony. The feud 
 between the fishing birds and their oppressors, the frigate birds, 
 is only active in the air; if the gannet or booby can but reach 
 the land and plant its feet on the ground the pursuer gives up the 
 chase immediately. 
 
 Beside the birds there were but few original inhabitants found 
 upon the islands. Among those I observe several varieties of 
 spiders, at least two of ants, a peculiar species of fly that at- 
 taches itself to the larger birds, and the common house fly, 
 which latter, however, may have been recentlj'' introduced. 
 They as well as common red ants are exceedingly abundant. 
 
 Rats were found on all these islands, especially on Howland's, 
 where they had become astonishingly numerous. It would seem 
 that they had been carried there long ago, as there are no traces 
 of recent shipwreck on the island, and had multiplied extensively. 
 On Jar vis' Island they were much less numerous, and were 
 probably brought by a ship that was wrecked there thirty years 
 since. They subsist on eggs, and also, as I observed on Baker's 
 Island, by sucking the blood of the smaller birds — the tern and 
 noddies ; and in this connection I may observe that these smaller 
 kinds of birds, described above, are almost entirely wanting on 
 Howland's, and their absence, I think, may be attributed to the 
 depredations of the rats. These rats of Howland's Island were 
 almost as numerous as the birds. They are of very small size, 
 being hardly larger than a large mouse, and, I think, must have 
 degenerated from their original state in consequence of the 
 change of climate, food and condition of life. They had com- 
 pleteh'' overrun the island, and on its first occupation by men 
 were a great annoyance. For many nights in succession a barrel 
 containing a few oats caught over 100, and I have known over 
 
18 J. D. Hague on the Guano Islands of the Pacific Ocean. 
 
 3,300 to have been killed in one day by a few men employed 
 for the purpose. 
 
 A species of small lizard was also found in great numbers on 
 Howland's Island, some specimens of which I had preserved in 
 spirit, but the package containing them was lost on the voyage 
 home. 
 
 Remains of former visitors. — There are some interesting traces 
 on this (Howland's) island of former visitors or residents. Ex- 
 cavations and mounds in the centre of the island, among the 
 thickets of brushwood, referred to above, are evidently the vv'ork 
 of man. The most extensive of these excavations is several 
 hundred feet long, and about one hundred feet wide, and ten or 
 fifteen feet deep, forming a gully or ditch, on each side of which 
 the sand and gravel is carefully banked up and kept in its place 
 by walls laid up of coral stone, (blocks of beach and reef rock). 
 
 The trees themselves may possibly owe their existence here 
 to the originators of these works, for the sides of this gully are 
 covered by «- growth of wood which, unless younger than the 
 rest, would show the trees to be of more recent origin than the 
 excavation. 
 
 It is said to be of a species called by the natives of the Sand- 
 wich Islands "Kou,"* which abounds on many islands of the 
 Pacific. In the same vicinity there are also the remains of what 
 were low, flat mounds of regular shape, formed of gravel and 
 walled up all around, being about a foot high, and just such as 
 I have observed are used by many South Sea Islanders for the 
 foundation and floor of their houses. In another part of the 
 island, near the western beach, some remains of a hut were found, 
 and near by the fragments of a canoe, some pieces of bamboo 
 and a blue bead. Here also was found, buried under a foot of 
 sand, a human skeleton, the greater part of which, on being ex- 
 posed to the air, crumbled to dust, leaving only two or three 
 bones in condition to be preserved. 
 
 On the south end of the island there is a foot-path laid to cross 
 a bed of coral debris or beach accumulations. The edges of the 
 corals being rough, sharp and painful to the feet, the paths 
 seems to have been laid for the convenience of passengers across 
 this end of the island. It is several hundred feet long, made of 
 flat, smooth stones, at convenient distances apart, for stepping 
 from one to the other. They were evidently laid by hand, as 
 they lie in a direction which forms nearly a right angle with the 
 ridges made by the sea. It is probable that the originators of 
 these works were South Sea Islanders. It sometimes happens 
 that they are drifted off to sea by currents in their canoes, and 
 such a party may have been thrown upon this island. No im- 
 plements or other traces of civilized people have been found. 
 
 ■ * Not to be confounded with " Koa," a species of acacia, and quite a different 
 tree. I have seen the Kou alluded to somewhere as a species of cordia. 
 
/, D. Hague on the Guano Islands of the Pacific Ocean, 19 
 
 It is not unlikely that the lizards which abound on the inhab- 
 ited islands of the Pacific were brought here by these people, 
 and the rats, possibl}^, came from the same source. 
 
 Other Islands. — As already observed, the discovery of these 
 deposits of guano, the extent and value of which were at first 
 greatly exaggerated, induced fortune-seeking parties to explore 
 the Pacific in the hope of finding many more of similar charac- 
 ter. Under the act of Congress of 1856, granting American pro- 
 tection to the discoverers and occupants, under certain conditions, 
 of such newly found deposits, nearly all the islands found on the 
 charts within ten degrees north or south of the equator and 
 within 150° and 180° W. were represented as possessing deposits 
 of guano, and claimed by parties who evidently knew but little 
 of their true condition. 
 
 A list, forty-eight in number, comprising nearly all of these 
 islands, was published in the New York Tribune, in March, 1859, 
 and was copied and discussed by Mr. E. Behm, in his interesting 
 and valuable article, entitled "Das Amerikanische Polynesien," 
 printed in Petermann's Mittheilungen, for 1859. 
 
 Of these islands, a number of which I have myself examined, 
 it is safe to assert that some, although having a place on the 
 charts, do not really exist, while many are of very doubtful ex- 
 istence ; in some cases two or more names are applied to the 
 same island; some are inhabited, others are covered with trees 
 and vegetation, and very few have guano on them.* 
 
 * The following is the list referred to. Those named in the first column are 
 islands whose existence and position is well authenticated, and the greater part of 
 which I have myself visited. Those in italics are either known or said to be guano 
 islands. Those marked with an asterisk are inhabited. Some of the islands men- 
 tioned in the second column are known certainly to exist, and are described by va- 
 rious navigators, while others are doubtful, but I am unable to add any positive in- 
 formation concerning any of them. The existence of those in the third column is 
 considered as highly improbable, at least in the position commonly assigned to them. 
 
 I. II. HI. 
 
 Baker's, 
 
 Caroline, 
 
 Danger, 
 
 (6^ 
 
 ' 30' N., 
 
 Arthur's 
 
 Jarvis', 
 
 Ann's, 
 
 162° 
 
 32' 
 
 W.), 
 
 Favorite, 
 
 Howkmd^s, 
 
 Staver's, 
 
 Makin, 
 
 
 
 Farmer's, 
 
 Maidens, 
 
 Flint, (10° 32' S., 
 
 Matthew's, 
 
 
 Sideron's, 
 
 Birnie's, 
 
 151° 05' W.) 
 
 Barber's. 
 
 
 Flint, (11° 26' S. 
 
 Phoenix's, 
 
 Baumann's, 
 
 
 
 
 162° 48' W.) 
 
 JEnderbury's, 
 
 Rogewein's, 
 
 
 
 
 Walker's, 
 
 Christmas, 
 
 Gronique, 
 
 
 
 
 Sarah Anne, 
 
 Clarence,* 
 
 Frieuhaven, 
 
 
 
 
 Samarang, 
 
 Duke of York,* 
 
 Quiro's, 
 
 
 
 
 David's. 
 
 Penrhyn's,* 
 
 Low, 
 
 
 
 
 
 Rierson's,* 
 
 Pescado, 
 
 
 
 
 
 Humphrey's,* 
 
 Ganges, 
 
 
 
 
 
 Danger,* (10° O'S., 
 
 Frances, 
 
 
 
 
 
 165° 56'" W.), 
 
 Mary Letitia's, 
 
 
 
 
 
 Palmyra, 
 
 Kemin's, 
 
 
 
 
 
 Sydney, 
 
 America, 
 
 
 
 
 
 Mary's, 
 
 Prospect, 
 
 
 
 
 
 Nassau. 
 
 
 
 
 
 
 Am. Jour. Sci. — Second Sekies, Vol. XXXH', No. 101. 
 
 -Sept., 1863. 
 
20 /. D. Hague on the Guano Islands of the Pacific Ocean. 
 
 i 
 
 The following named islands, in particular, have been sup- 
 posed, erroneously, as regards some of them, to have guano 
 deposits : 
 
 Latitude. Longitude. 
 
 McKean's, 3° 35' S., 174° 17' W. 
 
 Phcenix, 3° 40' S., 170° 52' W, 
 
 „ ) Enderbury's, 3° 08' S., 174° 14' W. 
 
 <^^ ( Birnie's, 3° 33' S., 171° 33' W. 
 
 Maiden's, 4° 15' S., 155° W. 
 
 Johnston's, 16° 53' N., 169° 31' W. 
 
 Christmas, 1° 53' N., 157° 32' W. 
 
 Starve, Starbuck or Hero, 5° 40' S., 155° 65' "W". 
 
 Of the above those of the Phoenix group are probably the 
 most important. McKean's Island has been occupied since 1858, 
 and several cargoes of guano of good quality have been brought 
 from it to this country. It is a low island, circular in form, not 
 exceeding three-fourths of a mile in diameter. Like Jarvis, it 
 once contained a lagoon though not elevated so high above the 
 sea. Its surface is consequently depressed, and is so much lower 
 than the beach that at high tides the guano deposit is sometimes 
 covered by two feet of water. As at Jarvis, a deposit of sulphate 
 of lime has resulted from the evaporation of sea water in the ba- 
 sin, forming the foundation on which the guano rests; and ow- 
 ing, probably, to frequent inundations, the two have become so 
 intimatelj' mixed that the quality of much of the guano is con- 
 siderably impaired. The better specimens contain about fifty 
 per cent phosjihate of lime mixed with much sulphate of lime. 
 Much of the deposit is covered by a foot of coral mud, which 
 has been spread out over the part adjacent to the beach. 
 
 Phoenix's Island is likewise very small, nearly circular, and 
 less than a half mile in diameter. The centre is considerably 
 lower than the beach, which is about eight or ten feet high, and 
 it is often flooded at high tides. I was unable to land on this 
 island, and my opportunities for observation were confined to 
 shipboard. The guano deposit cannot be very extensive though 
 said to be of good quality. 
 
 Enderbury's Island is described as an elevated lagoon, about 
 eighteen feet high, three miles long by two and a half broad. 
 It is said to contain deposits of guano, as is also its neighbor, 
 Birnie's Island, of which I am unable to give any positive in- 
 formation, having never visited either. 
 
 Maiden's is a large island, ten miles long, and said to be about 
 forty feet high. I believe it is an elevated lagoon, but much of 
 the guano deposit lies on the elevated ridge. Specimens which 
 I have examined, though free from sulphate, were much adulter- 
 ated by coral sand. Excepting McKean's, no cargoes have been 
 brought from these islands just alluded to. From Johnston's 
 Islands one or two cargoes have been brought to this country, 
 the greater part of which proved, I believe, to be sand. These 
 
/. D. Hague on the Guano Islands of the Pacific Ocean. 21 
 
 are described as three small islands (probably islets of one atoll) 
 containing but little guano and that much mixed with coral sand. 
 
 Christmas Island is a well-known lagoon thirty miles long, 
 trending east and west, having much vegetation. Much has 
 been said by speculators of its rich deposits, but I have good 
 reason to believe that there is no guano, worthy of mention, on 
 the island. Samples that I have examined were chiefly coral sand. 
 
 Starbuck's, Starve or Hero Island is an elevated lagoon, and 
 is worthy of mention because, like Jarvis', McKean's and other 
 islands of similar structure, it contains a large deposit of gypsum. 
 Its supposed guano I have found to consist of the hydrated sul- 
 phate of lime, containing about twelve per cent of phosphate of 
 lime and colored by a little organic matter. 
 
 So far as my observation extends, all elevated lagoons have 
 similar deposits of gypsum. 
 
 As regards the distribution of these phosphatic guano deposits 
 I believe them, in this region of the Pacific, to be confined to 
 latitudes very near the equator where rain is comparatively of 
 rare occurrence. In latitudes more remote from the equator 
 than 4° or 5° heavy rains are frequent, and this circumsiance is 
 not only directly unfavorable to the formation of guano deposits 
 but it encourages vegetation, and when an island is covered 
 with trees and bushes, the birds preferring to roost in them, 
 there is no opportunity for the accumulation of guano deposits. 
 
 New York, August, 1862. 
 
LIBRARY OF CONGRESS 
 
 D00275bH2Q3