8fcL 
 
 ^^>\ " .-^^k < 
 
 ■IK: 
 
Class S5S3 
 
 Dook > J (a 
 
 SMITHSONIAN DKl'OSIT. 
 
[from the AMERICAN JOURNAL OF SCIENCE, VOL. XXXII, SEPT. 1861.] 
 
 AGRICULTURAL CHEMISTRY-SOIL-ANALYSIS. 
 
 NOTICE OF THE AGRICULTURAL CHEMISTRY OF THE GEO- 
 LOGICAL SURVEYS OF KENTUCKY- AND ARKANSAS* 
 
 ~^^H> j7 BY PROF. S. W. JOHNSON. 
 
 >^^^j^ OF YALE COLLEO'te. 
 
 ^5, . 
 
 In no country has there been so much popular appreciation of 
 practical science as in the United States of America. Scarcely 
 one of the States is without its volume or volumes of Geological 
 and Natural History Reports, and though some of them have 
 been content to conhne the work to the merest outline of the 
 general and industrial geology of their territory, and have ex- 
 pended but a few hundi^eds of dollars in the undertaking, others, 
 like New York, have embraced all the branches of Natural Sci- 
 ence in their survey, have prolonged the work of exploration or 
 elaboration through many years, and have devoted money to 
 these objects with unsparing hand. 
 
 The results of these surveys as they stand recorded in the 
 numerous volumes published by the States and by the General 
 Government, are of very unequal merit, as might be expected 
 from the wide range of country explored, from the various degrees 
 of interest and appreciation governing the many Legislatures 
 which have authorized these, labors and from the exceedingly 
 unequal ability of the^dividuals charged with their execution. 
 
 These explorations have originated in all cases with our scien- 
 tific men. It is their influence either brought to^oear immediately 
 upon the legislative bodies, or exerted less directly through 
 cultivated and public spirited persons to whom the possible 
 benefits of geological surveys have been explained — that has 
 accomplished this vast work. 
 
 The enterprises of which we speak being sustained pecuniarily 
 at the expense of the people, and depending from year to year 
 in many cases upon the popular vote, it has been not only politic 
 but right to exhibit at the outset the prospects of pecuniary 
 return for the required outlay of means, as an inducement to 
 support such undertakings. It has been no less proper in pre- 
 senting the results of the surveys, to lay stress on the discoveries 
 having industrial bearings which are the fruits of the work. 
 
 In those States where large quantities of metallic ores occur, 
 the interest of capitalists engaged in mining has often sufiiced to 
 
 * 1st, 2d. Sd and 4th Reports of the Geological Survey of Kentucky ] 854-60: 
 2d Report of the Geological Reconnoisance of Arkansas, 1860: Agricultural Chem- 
 istry and Geology by Dr. D. D. Owen, principal Geologist, and Dr. Robert Pbteb, 
 Chemical Assistant. 
 Am. Joub. Sci.— Second Sbbies, Vol. XXXII, No. 95.— Sept., 1861. 
 30 
 
2 >S. W. Johnson on the Soil-analyses of the [234] 
 
 inaugurate a geological survey. In other states the agricultural 
 sentiment has had to be operated upon. 
 
 Great results have been promised to agriculture from the appli- 
 cations of geology and chemistry, and a great deal of labor has 
 been performed in the attempt to satisfy the hopes that have 
 been thus excited. 
 
 The chief object of the present notice is to inquire what has 
 been really accomplished for the good of the farmer, by the sci- 
 entific surveys that have been hitherto prosecuted in this country. 
 
 The labors of Dr. Peter in connection with the Kentucky and 
 Arkansas Surveys being the most recent and extended attempts 
 of this kind, we shall make them the basis of our inquiries. 
 
 If we except a few pages of general remarks on the theory of 
 vegetable nutrition, &c., which while useful to the practical readers 
 of the Keport contain no new facts or principles, — the whole 
 effort of Dr. Peter has been concentrated on the analysis of soils, 
 marls, rocks and ashes. He publishes in the four Kentucky Re- 
 ports analyses of 375 soils, and in the Arkansas Report, 187, in all 
 662 soil analyses. Besitles, we find the results of examinations 
 of 145 rocks, shales, &c., and of 38 ashes of plants, making a 
 grand total of 795 agricultural analyses. 
 
 The agricultural fruits of the surveys of Kentucky and Ar- 
 kansas are then to be sought in these analyses. 
 
 It certainly will strike all that the amount of work performed 
 by Dr. Peter is unusually great. It is now but six years since 
 the Kentucky survey was commenced and in that time the Dr. 
 has not only analyzed 795 soils, but has executed 516 analyses of 
 ores, slags, mineral waters and coals, making an average of two 
 analyses for every three days of this whole period. This labor 
 Dr. Peter states he has accomplished with the help of one intelli- 
 gent assistant, and by a special organization of his laboratory 
 and his operations whereby the utmost economy of time was 
 secured. We have had such experience of the advantages of a 
 similar system, that we are not prepared to doubt that the chemist 
 who adopts a plan of analysis which fully satisfies him, and from 
 which he never departs, may execute such an amount of work. 
 At the same time we must bear in mind that the only control 
 Dr. Peter offers for the accuracy of his results is, that the sum of 
 the weights of the separated ingredients equals their original 
 conjoined weight, no time being allowed to repeat a determina- 
 tion, or to prove the purity of a precipitate. 
 
 The Analytical Process followed in these analyses is not by 
 any means so minute and full as we should be warranted to 
 expect, when their author declares (4th Ky. Rep., p. 57) that 
 "such a work to be eminently useful must be thorough and 
 exhaustive ;" for soluble silica, chlorine, nitric acid and ammonia 
 are not at all estimated, and the condition of the iron, whether 
 protoxyd or peroxyd, is not noticed. It is worthy of notice that 
 
 b 
 
[235] Geological Surveys of Kentucky and Arkansas. 3 
 
 carbonic acid and lime are always present in atomic proportions 
 in the soils latterly analyzed, no excess of either ingredient being 
 mentioned in the results. Carbonic acid however is not noticed 
 in the description of the analytical process, and that figuring in 
 the analyses does not appear to have been directly estimated, 
 but to have come from the oxalic acid of the reagent shelf. 
 
 If, as might easily happen, the contrary not being proved, a 
 portion of the lime dissolved by hydrochloric acid exists in these 
 soils as silicate, sulphate or phosphate, then the assumption that 
 it is united to carbonic acid introduces an error into the summing 
 up (which in many cases is exactly 100) and shows that a quan- 
 tity of some other ingredient has been overlooked. 
 
 For the estimation of phosphoric acid a highly modified form 
 of Sonnenschein's process is employed, but our author does not 
 give the figures which prove that his changes are improvements. 
 
 Admitting however that the analyses are correct — we next 
 inquire what is their value — what useful deductions from them 
 appear in these Reports. 
 
 In the introduction to Vol. i, Kentucky Survey, page 13, Dr. 
 Owen says: " By consulting the numerous interesting results 
 obtained by the chemical analyses of the soils embodied in the 
 pages of this report, abundant evidence will be gathered of the 
 vital necessity of wide dissemination amongst the farming com- 
 munity, of the knowledge to be obtained by a correct insight 
 into their chemical constitution." In the same volume, page 
 373, Dr. Peter remarks that he was impressed " that when the 
 composition of our Kentucky soils and minerals in general, is 
 once accurately established, their applications to our wants and 
 uses would be obvious to all well informed persons. He has 
 therefore consumed the time mainly in the analyses, and made 
 up his report principally of the results." 
 
 In the agricultural section of the Arkansas Survey, p. 47, Dr. 
 Owen says: — "principally from chemical soil-analyses can the 
 agriculturist form an intelligent opinion as to the comparative 
 fertility of soils, and their suitability to the growth of certain 
 plants, as well as judge what applications may be required in the 
 way of lime, bone earth, plaster of Paris, ashes, or salts of pot- 
 ash, soda, &c." 
 
 Dr. Peter, in the same volume, page 166-7, observes: — "It is 
 believed that by no other mode than by chemical analysis or by 
 the more tedious and laborious method of actual experience, in 
 cropping for a series of years and publishing a record of the 
 same, can the actual nature, capabilities and value of the various 
 soils of a State be presented to the public ; and that by institu- 
 ting this Geologico- Agricultural Surve}', the State of Arkansas 
 not only aids materially in the progress of the general science 
 of the civilized world, and that of the soil in particular, but 
 takes the most effectual mode of making known to the enlight- 
 ened immigrant her agricultural riches. In this she has followed 
 
4 S. W. Johnson on the Soil-analyses of the [236] 
 
 the wise lead of the older state of Kentucky, in which, since 
 the institution of her geological surve}', the value of the land 
 in the regions examined and reported on has been very greatly 
 enhanced." 
 
 In the Agricultural Geology of Kentucky, Report 2d, p. 9, 
 Dr. Owen says: " Placing implicit reliance on the capabilities of 
 chemical science to indicate by the analyses of soils, the ingre- 
 dients removed by the cultivation and harvesting of successive 
 crops, it was hoped that by collecting samples of the virgin soil, 
 and of the same soil from an adjacent old field, that not only 
 the different substances assimilated out of the soil could be as- 
 certained, but also the exact proportion of these so that the far- 
 mer might know precisely what must be restored to the land to 
 bring back its original fertility." 
 
 These quotations sufficiently show what were the opinions 
 which led our author to devote such an amount of labor to the 
 analysis of soils, and indicate in general, what results were ex- 
 pected. 
 
 In the 2d Arkansas Report, p. 49 et seq., Dr. Owen " proceeds 
 to explain in what way soil-analysis becomes of value to the 
 farmer." He desires ^' to call particular attention to this subject, 
 because the opinion has been expressed even in this year (1860), 
 and by those having a high standing in the scientific world, that 
 chemistry is incapable of conveying any useful information to 
 the farmer by analyzing his soil." 
 
 On the six following pages of the 2d Ark. Rep., and on page 
 80 of the 4th Ky. Rep., Dr. Owen gives the most complete resume 
 of the teachings of soil-analysis which we are able to find in 
 the fi.ve volumes before us, and as these are his latest writings 
 on the subject, and as he then had the data of 389 analyses, viz. 
 of 187 Arkansas soils and 202 in the three volumes of the Ken- 
 tucky Report, — these being refrered to on the pages we are quoting 
 from, — we are warranted in considering what he has here pre- 
 sented, as embodying the strong points in favor of soil analysis. 
 We will notice them separately as gathered from both Reports. 
 
 1st. " Any one who will take the trouble to inspect the analy- 
 ses of the 187 Arkansas soils will see that the relative propor- 
 tions of the eleven mineral constituents of these soils is very 
 accurately given." — 2d Ark. Rej)., p. 49. 
 
 If we admit fully that Dr. Peter's analyses represent with 
 fair accuracy the composition of the two grammes of soil he ex- 
 perimented with in each instance, we do not therefore allow 
 that the composition of "these soils" considered as representing 
 geological formations, or large agricultural districts, or even 
 single fields, is "very accurately' given." 
 
 Here at the outset the distinguished gentlemen who have con- 
 ducted the 'geologico-' and 'chemico-agriculturai' part of the 
 Kentucky and Arkansas surveys have taken for granted, what 
 
[237] Geological Surveys of Kentucky and Arkansas. 5 
 
 being an error, overturns their whole reasoning, and renders 
 their soil-analyses comparatively worthless. 
 
 Years ago, following the teachers of agricultural chemistry in 
 this country and England, we believed that soil-analyses were 
 adapted to be of exceeding use to farmers. Having practised 
 analytical chemistry sufficiently to undertake the work, we pro- 
 ceeded, when on a vacation visit, to collect some farm soils for 
 the purpose of applying our skill and knowledge. On putting 
 down the spade and post-augur into the drift overlying the low- 
 est Silurian of Northern New York, we were at once struck 
 with the difficulty of procuring an average specimen. The soil 
 for a depth varying from two to six inches was quite fine, but 
 below that depth largely mixed with gravel. On comparing 
 different samples taken from a small area, it was plain that the 
 «oil was not a fit subject for analysis. The relative quantities 
 of organic matter as indicated by the color of the surface of small 
 stones, — some quartz and granitic, others slate and limestone 
 of several geological members, — were astonishingly variable. 
 Here we found the soil sandy, there it was clay. To take a 
 sample from one place was to do obvious injustice to the sixty- 
 acre field. To take it from a dozen places would not render the 
 selection of a fair sample any more certain. Then as to depth 
 — was it proper to go down six inches, one foot, or how far? 
 Had the field been a bed of iron ore, assays of a dozen samples 
 taken from different parts would have indicated very satisfacto- 
 rily the general value of the deposit, would have served as data 
 for buying and selling the property, because the worth of an 
 unworked bed of such ore depends less upon its content of iron 
 than upon external circumstances which affect the extracting of 
 the metal. Had the field been covered with rich dressed copper 
 ore to the depth of six inches, it would have been necessary to 
 divide it up into small parcels of a few tons, average these care- 
 fully and as carefully assay each one. No one would risk pur- 
 chasing a hundred thousand tons of copper ore on the analysis 
 of one or of a dozen samples, because it is impracticable to in- 
 termix or average such a mass of material as that a dozen sam- 
 ples shall accurately represent it. 
 
 We hold it therefore as the first objection to soil-analyses that 
 to procure a specimen which accurately and certainly represent a 
 field or district, is practically impossible in the majority of cases, 
 and if possible^ requires a series of analyses to prove the fact. 
 
 This argument applies with the greater force when we con- 
 sider how small a proportion of the ingredients of a soil are of 
 any immediate use in feeding crops. The really active nutrient 
 matters of a soil are not reckoned by per cents nor by tenths of 
 per cents, but by the minutest fractions, 
 
 A heavy crop of thirtj^-seven bushels of wheat, grain and 
 straw included, removes from an acre of land but 300 lbs. total 
 
6 'S'. W. Johnson on the Soit-analyses of the [238] 
 
 of mineral matters. According to Dr. Peter's weighings on some 
 of the Kentucky soils, we may assume, that taken to the depth 
 of a foot, an acre of soil weighs 3,000,000 lbs. All that is re- 
 moved by the heaviest wheat crop then in one year is but ^-^ lo-o, 
 or 0-0038+ per cent. 
 
 It follows that the annual removal of the heaviest crop of 
 wheat from a soil for 100 years diminishes its mineral matters 
 by less than 04 per cent. If then, in the selection of a sample, 
 the aveVage composition is departed from to the amount of 4 
 parts in 1000, the analysis may misrepresent the soil, by the 
 value of 3700 bushels of wheat per acre, or by what represents, 
 so far as mineral ingredients can, the fertility of a century. 
 
 What freaks and accidents is not the soil-analyst the sport of? 
 A bird, squirrel, or dog relieving nature at the spot where he 
 collects his sample, innocently magnifies the phosphoric acid or 
 alkalies of the surrounding thousand acres a hundred fold. 
 The soil gathered toward the end of a long rain, whereby its 
 soluble matters are carried deep into the subsoil, is declared poor, 
 by analysis, whereas if taken after a fortnight of drought it 
 might appear extraordinarily fertile. Boussingauit found in his 
 rich garden soil in June, during wet weather, 00003-1 per cent 
 of nitric acid. In the following September, after a period of 
 dryness, it contained 0*0093 per cent, or twenty-seven times as 
 much as in June. This ingredient is indeed more liable to fluc- 
 tuation in amount than any other, both because it is formed in 
 the soil, and because it is not subject to the absorbent action 
 which the soil exercises over most other of its soluble constitu- 
 ents ; but the same variation occurs among the other ingredients 
 according to the direction of the capillary movement of the soil- 
 water, though in less degree. 
 
 Independently however of all considerations and calculations 
 like the above, we have proof — evidence at least that supports 
 these considerations, and has never been publicly refuted — that 
 it is practically impossible to obtain average specimens of the 
 soil. I refer to investigations made as long ago as 1846-9 under 
 the direction of the Prussian " Landes Oekonomie Collegium,^'' 
 and reported by the distinguished Magnus. The account of 
 these experiments is given in detail in Erdmann's Journal fiir 
 Praktische Chemie, vol. xlviii, pp. 447 et seq. 
 
 The ^''Landes Oekonomie CoUegiian" at that time carried on 
 systematic experiments in agriculture at fourteen distinct stations 
 scattered through the Prussian domain. The trials which we 
 now speak of, were made for the ostensible purpose of studying 
 the exhaustion of the soil by cropping. The plan was to analyze 
 the fourteen soils, the history of which for years previous was 
 accurately known, then crop them with rape until "exhausted," 
 then compare together the original composition of the soils with 
 their composition after exhaustion, taking into account as well, 
 
[239] Geological Surveys of KentucJcy and Arkansas. 7 
 
 the composition of the crops removed. The research began 
 with collecting and analyzing the soils. In order to meet as far 
 as possible the difficulties of securing average specimens, equal 
 portions of the soil of each field were taken with the spade at 
 ten or twelve different points, and thoroughly intermixed ; each 
 sample was then passed through a sieve, the holes of which were 
 two square lines in area, so as to remove all coarser stones, then 
 again well worked over to complete the mixture. Of each sam- 
 ple three separate portions were analyzed, in most cases by dif- 
 ferent operators. The analyses were made by, or under the 
 guidance of, the ablest chemists of Germany. They were made 
 according to a prescribed scheme, and that there should be no 
 reason to slight the work, the work was paid for. It is true that 
 analytical chemistry was not so advanced in 1846 as now. It is 
 true that the methods then practised for estimating phosphoric 
 acid and some other substances were not as perfect as they now 
 are ; but for the most part the analyses then made are as accu- 
 rate as any that could be executed to-day. It cannot be sup- 
 posed for a moment that analysts like Eammelsberg, Bodecker, 
 Genth, Debus, Knop, Heintz, Krocker, Marchand, VV^eidenbusch, 
 Sonnenschein, Varrentrap, Weber, &c. &c., would by fault of 
 method or by carelessness return anything but results that were 
 accurate, as far as it was possible to make them such. We cannot 
 suppose that their determinations of lime, oxyd of iron, potash 
 and sulphuric acid, substances estimated then by the same meth- 
 ods that are now employed, would vary if they were supplied with 
 homogeneous material to operate upon. But let us look at some 
 of their figures. We tabulate a number of them taken at ran- 
 
 dom: 
 
 
 
 
 
 
 Soil of 
 
 Eldena, 
 
 !? 
 
 Lime. 
 
 0-39 
 0-75 
 0-25 
 
 Potash. 
 
 0-93 
 2-06 
 0-12 
 
 Sulphuric acid. 
 0-08 
 
 Phosphoric acid. 
 
 0-06 
 0-17 
 
 0-40 
 
 0-02 
 
 
 fa. 
 
 - b. 
 
 c. 
 
 0-802 
 0039 
 0-715 
 
 3-825 
 0-490 
 0-792 
 
 
 0-042 
 0-046 
 0-007 
 
 Beesdau, 
 
 
 0-004 
 
 Neuensund, 
 
 t 
 
 1-692 
 0-614 
 0-728 
 
 3-531 
 1-289 
 1-243 
 
 0-050 
 0-038 
 0-241 
 
 0-051 
 0-010 
 0-121 
 
 Turwe, 
 
 \l 
 
 2312 
 
 2-67 
 
 3-391 
 
 1-112 
 
 1-14 
 
 0-201 
 
 0-040 
 
 0-03 
 
 0-022 
 
 0-057 
 
 0-20 
 
 0-014 
 
 Frankenfelde, 
 
 \l 
 
 0-420 
 1-081 
 0-461 
 
 1-155 
 
 0-016 
 
 0-004 
 0-418 
 0-071 
 
 1-456 
 
 0-015 
 
 If we run over these figures and those of the entire series of 
 analyses, we find that different determinations disagree to such 
 an extent as to make it the sheerest folly to base any calculation 
 
8 S. W. Johnson on the Soil-analyses of the [240] 
 
 of the value of the soil upon analysis. Some of the analyses 
 agree sufficiently to show that accordant results are possible if 
 uniform material be taken ; but the grand result of the investi- 
 gation is that the difficulties of getting a uniform material are 
 exceedingly great. Again, we must remember that in the case 
 before us, the three examinations of each soil were made upon 
 portions of one carefully mixed sample. What would have 
 been the result had each chemist received a sample collected 
 separately from all the others, and from different parts of the 
 field ! 
 
 Dr. Peter mentions these analyses of the Landes Collegium, 
 and quotes a few of the results on page 187 of the 3d Kentucky 
 Beport. He believes however that these discordant results do 
 not invalidate soil analyses when made as they may be made 
 with " means and appliances now at the service of the analytical 
 chemist" and thinks "this statement however hazardous it may 
 seem will be found to be sustained" in his Report. 
 
 In the Report before us however we do not lind anything to 
 sustain Dr. Peter's view. He gives, so far as we have discovered, 
 no duplicate analyses, to show what accuracy his methods admit 
 of on the same sample, much less does he prove by analyses of 
 specimens separately gathered from the same field, that it is easy 
 to procure an average material for analysis. Until this proof is 
 produced the evidence is in favor of our view. 
 
 Having shown how small an error in sampling may affect the 
 chemist's estimate of a soil, it is not out of place to insist for a 
 moment, that a similar error in the analysis itself, must have the 
 same result. In running over 200 pages of Dr. Peters 4th Ken- 
 tucky Report, we find five analyses of soil in which there is a 
 gain of from five to eight tenths per cent; we find twenty -three in 
 which there is a loss exceeding five tenths per cent. In thirteen 
 of the latter the loss is eight or more tenths, in eight instances 
 the loss is one per cent or more, and in one case is one and eight- 
 tenths per cent. We should scorn to notice little matters like 
 these, errors which are inseparable from the best manipulation 
 and the best processes, were it not that in soil analysis it is pre- 
 cisely the small quantities which alone have any importance. 
 
 We find in Dr. Peter's work, as in the work of all who have 
 preceded him in the analysis of soils from Davy and Sprengel 
 down, evidence that the best endeavors in this line of research 
 are entirely incommensurate with the desired results. 
 
 It may be objected to this criticism of the analyses that the 
 loss or gain must be distributed among the twelve ingredients 
 determined. It is true that there is a probability that such dis- 
 tribution would be just; but this is by no means certain^ and it 
 is equally true that this being done there is still force in the 
 criticism — for the four-tenths per cent of the soil which a cen- 
 
[241] Geological Surveys of Kentucky and Arkansas. 9 
 
 tury of wheat crops would remove, likewise consists of twelve 
 ingredients. 
 
 The 2d result of these analyses, according to Drs. Owen and 
 Peter, is what the former (4th Ky. Eep., p. 33) declares to_ be 
 "a general law" " now established," viz., "that soil-analysis is 
 capable of showing the exhaustion in land of the mineral food of 
 plants by continual cropping.'''' 
 
 To show the removal of soil-ingredients by cropping, the 
 plan was followed of collecting soils from contiguous fields, one 
 of which had been " cultivated" while the other was in its virgin 
 state. On comparing the analyses it was found that in seventy- 
 one* cases out of seventy-nine, a loss had occurred in the 
 soil which had been in use without manure from ten to fifty 
 years. In eight instances, however, the analysis failed to show 
 such a result, owing to local causes, the soil of the old field 
 being based on a sub-soil richer than was the virgin field, or the 
 old field having received washings of more elevated lands, &c. 
 
 The admitted richness of the old over the new soil in these 
 eight exceptional cases, is expressed by hundredths of per cent, 
 e.g., soil Nos. 982, virgin, and 983, cultivated, differ by 0'066 per 
 cent of potash. Soils 1144 and 1146 by 0-032 per cent of phos- 
 phoric acid. Soils 1204 and 1205 by 0*092 per cent phosphoric 
 acid. Soils 1207 and 1208 by 033 per cent potash. Similar 
 fractions likewise show the amount of deterioration in the other 
 seventy -one cases. We adduce two instances pointed out by 
 Dr. Peter in the 3d Kentucky Keport, p. 207, and one given on 
 p. 176 of the 2d Arkansas Peport : 
 
 Carb. of lime. Magnesia. Phosphoric acid. Potash. 
 
 Virgin soil, No. 557, 0-345 0-335 O'lSlf 0-156 
 
 Old soil, No. 558, 0-215 0-465 O-lOSf 0-101 
 
 Difference, 0-130 0-130 gain. 0-078f 0-055 
 
 Virgin soil, No. 738, 0-180 0-444 0-179 0256 
 
 Old field, No. 739, 0-145 0-388 0-163 0-179 
 
 Difference, 0-035 0-056 0016 0077 
 
 Virgin soil. No. 288, 0-121 0-371 0-127 O'llS 
 
 Old field, No. 289, 0-021 0-371 0-053 0-097 
 
 Difference, 0-100 0000 0'074 0-019 
 
 We were prepared to find these differences much larger. It 
 is seen at a glance that they fall within the errors of Dr. Peter's 
 own manipulation, and when we assert that of ten analyses of 
 the most homogeneous material made by the same analyst under 
 the most favorable circumstances, five would differ among each 
 other by an amount equal to the quantities upon which this 
 " natural law " is supported, we assert what every competent 
 
 * Misprinted twenty-one, on p. 31, 4th Kentucky Report. 
 
 f Misprinted on p. 207, 3d Ky. Rep, wliere the difference is made 0-045 instead 
 of 078, as given above frona the tabulated analyses. 
 Am. Jour. Sci.— Second Series, Vol. XXXII, No. 95.— Sept., 1861. 
 31 
 
10 S. W. Johnson on the Soil-analyses of the [242] 
 
 analyst knows to be true, and what moreover pronounces most 
 emphatically upon the value of such investigations. 
 
 It is therefore our conclusion, that while, as has long been 
 known, the soil loses in mineral matter what the crop gains, it is 
 doubtful if in any given case chemical analysis can indicate this 
 difference with certainty, for the reasons that the accidents which 
 affect analysis make the limits of inaccuracy, to cover more than 
 the loss by years of cropping. When we take into account the 
 changes that are constantly progressing in the soil when under 
 cultivation — changes by which the disintegration is hastened, 
 changes by which it is made in many instances more retentive 
 of soluble matters — when we remember that most cultivated 
 crops, although they carry oif in seed, stem and foliage a quan- 
 tity of mineral matters, yet derive these in part from a depth 
 below the range of analysis, and in their roots or stubble, leave 
 upon the surface, salts brought up from a considerable depth — 
 we perceive that the problem is so complicated with compensa- 
 tions and variable quantities as to put it beyond the reach of 
 quantitative chemical analysis. 
 
 If, in any case, soil-analysis does show or appear to show the 
 exhaustion of the soil, it is however, the appeal to experience 
 which proves it, and as this is the first, most obvious, and an 
 entirely sufficient proof, we do not see the value of the " law " 
 that has 10 per cent (eight-seventy-ninths) of exceptions, the 
 existence of which like that of the rule itself, is only to be 
 established by comparison with the plain agricultural fact. 
 
 In short, if we admit the result as Drs. Owen and Peter would 
 have it — of what use or interest is it? 
 
 The 3d point, is that analysis shows " the peculiarities of the 
 soils derived from different geological formations." Says Dr. 
 Owen, "these analyses most distinctly show that certain geolog- 
 ical formations impart to the soil more of the important mine- 
 ral fertilizers than others." The reader will be able "to see 
 that it is those formations which are composed of easily disin- 
 tegrating materials, which, all other things being equal, yield the 
 soils richest in phosphoric acid, lime and potash; and at the 
 same time contain the quantity of alumina and oxyd of iron 
 necessary to render them sufficiently retentive and attractive of 
 atmospheric water and ammonia; therefore these soils are the 
 best adapted for those grains and crops which require the largest 
 proportion of these ingredients." "He will moreover be able 
 to trace the gradual diminution in the proportion of the more 
 important mineral ingredients, down from these extraordinarily 
 fertile soils derived from the highly fossiliferous, argillo-calca- 
 reous beds of the lower Silurian, the Cretaceous and the Tertiary 
 systems of the West; through the silico-calcareous soils of the 
 upper Silurian, Devonian and Sub-Carboniferous limestone strata, 
 in which fossils are either more sparingly distributed or, in some 
 
[243] Geological Surveys of Kentucky and Arkansas. 1 1 
 
 cases almost wanting, and which are far less easy of decomposi- 
 tion; thence through the argillo-silicious soils of the Coal meas- 
 ures with only locally organic remains, and these chiefly of plants, 
 down to the more purely silicious soils prevalent where the 
 non-fossiliferous sandstones of the Coal measures and of the 
 Millstone Grit, prevail to the exclusion of either shales or lime- 
 stones and wfiich afford the most unproductive soils as yet an- 
 alyzed." While it is to be expected that rocks of complex origin 
 rich in organic remains — which are evidences that the rocks them- 
 selves originally resulted from the deposition of the washings of 
 fertile lands — should yield richer soils than sandstones or lime- 
 stones, w^e do not see that analysis of the soil makes the fact 
 more evident Knowledge of the composition of a rock enables 
 us to judge in a general way of the value of the soil, so far as 
 this depends upon chemical characters. We do not see what is 
 gained by further analyses of the soil. It would appear that the 
 cheap mental processes of deduction or inference may accomplish 
 here in a moment all that an expensive analysis can show. 
 
 We fail moreover to perceive that analysis shows " the pecu- 
 liarities of the soils derived from the different geological forma- 
 tions." In a cretaceous or limestone soil we of course expect to 
 find much carbonate of lime, and in a sandstone or millstone grit 
 soil much insoluble silica or silicates, but the quantities of phos- 
 phoric acid, potash and sulphuric acid do not appear to bear any 
 definite relation to their geological origin. It is impossible to 
 represent the composition of the soil of any geological formation 
 by a typical statement of percentages, or to point out its pecu- 
 liarities further than by an undefinable more or less. Although 
 Kentucky and Arkansas lie mostly or altogether beyond the 
 influence of drift, yet the action of running water in its con- 
 stant passage from hill-top to valley has to a great degree oblite- 
 rated from the soils those peculiar differences to be found among 
 the rocks from which they have been derived. 
 
 A careful examination of the analyses recorded in the Arkan- 
 sas survey shows that the average composition of the eight soils 
 analyzed from the Lower Silurian and of the fourteen from the 
 millstone grit, compare as follows, in regard to the more im- 
 portant ingredients : 
 
 Phosplioric Sulphuric 
 
 Carb. lime. Ma;^nesia. acid. acid. Potash. 
 
 Lower Silurian, average of 8 soils, 0-533 0-485 0184 0-052 0-355 
 
 Millstone grit, " 14 " 0-215 0531 0180 0-057 0-148 
 
 Here we see that the soils of the poorest formation are inferior 
 to those of the richest only in carbonate of lime and potash. Of 
 the soils of the millstone grit, nine are richer in carb. lime than 
 the poorest of the Silurian, and five of the former contain more 
 potash than the poorest of the latter. On the other hand but 
 two of the Silurian soils have higher percentages of either carb. 
 
12 -S. W. Johnson on the Soil-analyses of the [244] 
 
 lime or potash, than the richest soil of the millstone grit. If 
 these figures demonstrate anything, it is the fact, that no geolog- 
 ical formation has the absolute monopoly of either barren or 
 fertile soils. If the analyses of Dr. Peter show the "peculiari- 
 ties" of the soils of any geological age, then certainly these 
 peculiarities are not remarkably peculiar! 
 
 On page 50 of the 2cl Ark. Rep., Dr. Owen remarks as follows : 
 
 " With the table of the composition of the ashes of plants to 
 refer to, appended to this Report, and after becoming acquainted 
 with the usual proportions of mineral constituents in an average 
 soil, information which is easily acquired by looking over the 
 table of soil analyses in this Report, it is easy for any individual 
 to see, when he is provided with a reliable analysis of his soil, 
 not only to what crop it is best adapted, but what kind of min- 
 eral fertilizers, if any, it requires as a manure, and how it com- 
 pares in fertility to the various grades of soils from other farms 
 and other states. Is not this knowledge of some value to the 
 farmer?" 
 
 The above, we are of opinion, proceeded rather from the 
 generous heart than from the critical brain of its lamented author. 
 Had he attempted to do the things which he believed to be so 
 easy, we are sure his statements would have lost somewhat of 
 their directness and would have appeared in a form highly modi- 
 fied from the above. " The usual proportion of ingredients in an 
 average soil." What is an average soil? Our only way of 
 deciding what is such a soil consists in noting the average yield 
 of soils. But the yield depends not alone on the soil, but upon 
 climate, weather, tillage and various incidents and accidents. It 
 depends not on the composition of the soil — not on the "propor- 
 tion of ingredients" alone, but likewise on the condition of 
 those ingredients, their state of combination, their solubility. It 
 depends also on the physical characters of the soil, which deter- 
 mine the relations of the crop to the essential conditions of reg- 
 ulated heat and moisture. The soil is not less important to the 
 plant in its function of home than in its function of food, the lodg- 
 ings are of equal influence with the board. It is a nice work to 
 balance these varying circumstances, many of which have as yet 
 in our science, no shadow of a numerical expression, and then to 
 say how many thousandths of a per cent of potash, lime, phos- 
 phoric acid, &c., belong to the "average soil." 
 
 Dr. Peter has indeed attempted to show the degree of availa- 
 bility of the elements of the soil by the following process. "A 
 quantity, generally thirty grammes of the air-dried soil is 
 placed in an eiglit-ounce strong vial, with a close fitting stopper, 
 and the bottle is filled up with distilled water which has been 
 charged with pure carbonic acid gas, under a pressure of about 
 two atmospheres. The bottle is allowed to remain for about a 
 
[245] Geological Surveys of Kentucky and Arkansas. 13 
 
 month at a temperature about that of summer heat." The 
 matters thus dissolved were then analyzed as usual. These 
 results have this value, they show that the water of the soil is 
 capable of dissolving all the elements of the food of plants. 
 They furnish moreover a rough comparative view of the availa- 
 ble matters in different soils. Beyond this we cannot attach any 
 value to them. 
 
 We now come to Dr. Owen's 4th result of soil-analyses, 
 embodied in the above quotation, and repeated on p. 30 of the 
 4th Ky. Kep., viz: its power of indicating "the suitabilit}^ of 
 the soil for any particular crop." Closely related to this is the 5th 
 item, viz., that anaWsis can show " what addition any soil, either 
 uncultivated or cultivated, requires to render it productive and 
 remunerative for any given crop; and, of course, the deficiency 
 in the soil of one or more of the eleven elements determined by 
 chemical analysis." 
 
 We cannot help feeling that the above assertions which are 
 here made unqualifiedly, were intended to be understood with a 
 large amount of reserve and subject to various conditions. Oth- 
 erwise we must regard them quite unjustified, if not absurd. 
 The chemical analysis of soil reveals nothing as to its tenacity or 
 lightness, its porosity or retentiveness for water, j9t these phys- 
 ical and mechanical conditions more than anything else determine 
 the adaptation of a soil for any particular crop. The best grass 
 lands are not the best wheat lands — and although it would 
 scarcely be questioned that wheat requires a richer soil than 
 grass in order to produce an average crop, and although as we 
 know, it often happens that many successive hay crops may be 
 removed from a meadow without sensible diminution of the 
 yield, while uninterrupted cropping with wheat nearly always 
 reduces the capacity of the soil in a very few years below a 
 profitable point; yet each average hay crop removes from a field 
 more of every ingredient of vegetation than the grain and straw 
 together of an average harvest of wheat. 
 
 Such at least is the testimony borne by the most recent and 
 trustworthy data. Dr. Anderson of Glasgow basing his calcula- 
 tions on the best analyses and on the extensive agricultural 
 statistics gathered in late years by the Highland and Ag. Society 
 of Scotland, makes the following estimate of the amount of the 
 principal ingredients removed from an acre by average crops of 
 seven staple British farm products. See table on next page. — 
 Trans. Highland and Ag. Soc.., 1861, p. 568, 
 
 On comparing the amount of matters removed from an acre 
 by the wheat and hay crops, we find that the latter requires four 
 times as much potash, lime and sulphuric acid ; twice as much 
 silica and one-fifth more nitrogen. 
 
 Again we know that oats are raised on soils which are consid- 
 ered too poor for the profitable production of wheat, and the 
 
14 
 
 iS. W. Johnson on the Soil-analyses of the 
 
 [246] 
 
 s = 
 
 o CO 
 
 CO 
 03 
 
 00 CO 
 
 05 O 
 
 o 
 
 ^ o 
 
 lO r-( 
 
 CD 
 
 O O) 
 
 r-< GO 
 
 00 
 
 CO 
 
 o 
 o 
 
 
 ._ 0) 
 
 05 to 
 <N l-H 
 
 
 ^ CO 
 
 CO 
 
 Tj* 
 
 
 Tl^ 
 
 CO CO 
 
 04 
 
 o 
 CO 
 
 CO 
 
 CD 
 
 
 M 
 
 CO t— 
 
 O 
 
 OS 
 
 Oi 00 
 
 
 
 o\ 
 
 (M CO 
 
 00 
 
 o 
 
 o 
 
 OS 1— 
 
 C4 CO 
 
 EC 
 
 O CO 
 
 CO 
 
 CO 00 
 i-i CO 
 
 i-H 
 
 00 
 
 
 
 O CO 
 
 Tf* 
 
 CO 
 
 l-H CO 
 
 CO 
 
 l-H 
 
 11? 
 
 C^ CO 
 
 <N r-l 
 
 00 
 
 CO 
 
 
 
 O CO 
 
 
 CO r- 
 
 OS 
 
 CO 
 
 
 O OS 
 
 CO CO 
 I— ( 
 
 OS 
 
 t— 1 
 
 
 CO i>- 
 
 CO 
 CO 
 
 CO o 
 
 l-H 
 
 r-H 
 
 
 «0 '-' 
 
 — 'C-d 
 
 CM lO 
 
 CO »Ci 
 
 00 
 
 iC O 
 OO —1 
 
 05 
 
 • 00 
 
 • F-H 
 
 00 
 
 CD l^ 
 
 CO 
 
 00 
 
 O 
 
 
 ^IS 
 
 O CO 
 
 CO 
 
 O .-1 
 
 l-H 
 
 ! "5 
 
 lO 
 
 l-H <N 
 
 CO 
 
 OS 
 CO 
 
 O CO 
 
 c 
 
 • ; 
 
 • 
 
 (M O 
 
 r-i CO 
 
 0^ 
 
 o o 
 
 
 • O 
 
 • O 
 
 o 
 o 
 
 CO 
 
 l-H l-H 
 
 oi CO 
 
 
 • * 
 
 
 l-H l-H 
 
 (N 
 
 o »o 
 
 lO 
 
 \ c^ 
 
 I> 
 
 o 
 
 r-H 
 
 CO OS 
 CO 
 
 ^■i 
 
 O CO 
 00 CN 
 
 CO 
 
 o 
 
 o »o 
 
 l-H t^ 
 
 00 
 
 O iCi 
 
 
 O ^ 
 
 
 CO 
 
 l-H <N 
 
 r-< CO 
 
 ss 
 
 -* Ol 
 
 £- 
 
 ec <N 
 
 ^ 
 
 "* iCi 
 
 OS 
 
 '^ CO 
 
 r^ 
 
 -* 
 
 c^ OS 
 
 a 
 
 ^ CO 
 
 o o 
 
 t- 
 
 o 
 
 CO r-, 
 
 i^ 00 
 
 UO 
 
 l-H C5 
 
 CO <M 
 
 o 
 CO 
 
 l-H t^ 
 
 o CO 
 
 00 
 
 CO 
 
 o 
 
 CD 
 
 O CO 
 
 r^ CD 
 
 05 
 
 O O 
 
 CO 
 
 O O 
 
 o 
 
 CO OS 
 
 CO 
 
 00 
 
 l-H >0 
 
 CO 
 
 <i 
 
 o ^ 
 
 (N OQt 
 
 o 
 
 o -* 
 
 C.-> r-H 
 
 
 • o 
 
 o 
 
 i-H 
 
 l-H ^ 
 CO r-H 
 
 
 l-H 
 
 »o o 
 
 00 00 
 
 o 
 OQ 
 
 l-H (M 
 
 ■<i< 
 
 O 1-1 
 
 pH 
 
 '. ^ 
 
 CD 
 
 O CO 
 
 l-H 
 
 CO 
 
 
 CN ->*< 
 
 o o 
 
 r-H C<l 
 
 o -* 
 
 — "3 2 
 
 (M 00 
 
 l-H tJ* 
 
 o 
 CO 
 
 C^ l-H 
 
 CO 
 
 CO 
 
 O CO 
 00 lO 
 
 
 i^ r-l 
 
 OS >c 
 
 00 
 
 
 00 -H 
 
 iQ CO 
 
 
 CO ^ 
 
 l-H 
 
 00 
 
 l-H 
 
 -1* CJS 
 
 -<* OS 
 
 CO 
 
 00 CO 
 
 OS 
 
 l-H 
 
 in 00 
 
 lO O 
 
 CD 
 
 CO 
 
 IC l-H 
 
 in OS 
 CO 
 
 o o 
 
 CD (M 
 I-H (M 
 
 
 ^ 
 
 
 -D 
 
 
 ^ 
 
 
 ^ 
 
 
 
 
 '-"' 
 
 
 •~~» 
 
 
 r-^ 
 
 1 
 
 o 
 
 CD 
 
 -i-i 
 
 CO 
 
 
 o 
 
 
 o 
 
 CD 
 
 
 
 
 
 
 
 
 
 
 C3 
 
 CO 
 
 rt 
 
 cS 
 
 o 
 
 ca § 
 
 d: a 
 
 Iri 
 
 G 
 O 
 
 "Iri 
 
 00 
 
 -5 
 
 I-H 
 
 '« '-' 
 
 1— I 
 
 D 
 
 P 
 
 ^ 
 
 
 s 
 
 « 
 
 ^ 
 
 l-H 
 
 rO 
 
 
 r^ 
 
 
 ^ 
 
 
 CO 
 
 
 CO 
 
 
 ~* 
 
 
 lO 
 
 
 <N 
 
 
 CO 
 
 
 CO 
 
 
 (N 
 
 CO CO 
 
 •HJN 
 
 'heat- 
 Grain, 
 Straw 
 
 o 
 H 
 
 irley — 
 Grain, 
 Straw 
 
 C8 
 
 ats— 
 Grain, 
 Straw 
 
 15 
 o 
 
 ^ 
 
 
 PQ 
 
 
 O 
 
 
 p-i 
 
 C3 
 
 3 O cS 
 
 H ^ a 
 
[247] Geological Surveys of Kentucky and Arkansas. 15 
 
 table shows us that an average crop of oats requires more of 
 every mineral ingredient than is needful for a corresponding 
 wheat crop. 
 
 In fact, wheat is the crop, to grow which continuously requires, 
 according to universal agricultural experience, land richer than 
 that needed for any other of the seven crops whose chemical 
 statistics are given in Dr. Anderson's Table, and notwithstanding, 
 with exception of barley and the potato-tuber, it removes the 
 least from the soil. 
 
 The farmer knows that wheat delights in a deep, rather heavy 
 soil, one which holds moisture well, and yet is not wet. Barley 
 and oats flourish on soils that are too dry and light, and grass 
 on those which are too wet for wheat. 
 
 But how does the matter stand when these external conditions 
 are taken into account? Does not analysis aid us then in a 
 good degree? Let us take a case similar to what has repeatedly 
 occurred in actual practice. We have a soil which as the result 
 of long cultivation or from natural deficiencies, is incapable of 
 yielding a remunerative crop of wheat. Its texture is good, it 
 has produced wheat abundantly, and needs nothing but a little 
 of the right kind of manure to restore its power of giving a 
 crop. We put upon it Peruvian guano at the rate of 300 lbs. 
 per acre, and the harvest is a good one. The entire addition to 
 the soil is but gnlfsTToths = one hundredth per cent. The 
 amounts of phosphoric acid, of alkaline earths and nitrogen 
 added, are for each, but one six-hundredth per cent of the soil, 
 taken to the depth of a foot. These quantities are rather minute 
 for even the improved analysis of the present time to estimate 
 successfully. 
 
 Calculations like this show that the chemist cannot discrimi- 
 nate by his analysis between; 1st, a soil which is unproductive 
 from the temporary exhaustion of some of its available ingre- 
 dients ; 2d, the same soil which is rendered fertile again for a 
 year by the use of 300 lbs. of guano ; and 3d, the same, made 
 over-rich so that nothing will grow on it, by an application of a 
 ton of guano. 
 
 On page 18 of the 2d Ky. Eep., Dr. Owen remarks as follows: 
 "During last summer a soil was collected in Bullit county, from 
 an old field which had been fifty or sixty years in cultivation, 
 and which will now no longer produce clover. I venture to 
 predict that when the analysis of this soil shall be completed it 
 will be found to be deficient in some of these constituents,* and 
 the analysis will probably show what other green crop might 
 succeed better for the renovation of such land." 
 
 On page 230 of the 3d Ky. Eep., Dr. Peter gives the analysis 
 of this soil, and says, " The inability of this soil to produce clover 
 
 * Tlie mineral ingredients of plants. 
 
16 S. W. Johnson on the Soil-analyses of the [248] 
 
 is explained by its very small proportion of lime, and rather 
 small amount of sulphuric and phosplioric acids. The addition 
 of plaster of Paris or some of the calcareous marls would prob- 
 ably restore it to the capability of supporting a clover crop." 
 
 The percentages of the ingredients which Dr. Peter considers 
 deficient, are as follows : 
 
 Carbonate of lime, - - - 0-0'72 =lime 0-040 
 Sulphuric acid, ... 0-055 
 
 Phosphoric acid, . . . 0"070 
 
 Small as are these quantities, the smallest of them, viz., that 
 of lime, yet amounts to 1200 lbs. per acre, which is enough to 
 supply 10 clover crops of 3 tons each, and as by the analysis it 
 all exists in the form of carbonate, it must all be available. We 
 know from the vegetation experiments of Boussingault, Ville, 
 and Sachs, that plants are capable of absorbing from a limited 
 amount of soil the whole of any soluble nutritive substance pres- 
 ent, provided its quantity be no more than the plants require, 
 and the other elements of fertility are at hand in excess. 
 
 Twelve or thirteen years ago. Dr. Anderson in his capacity of 
 Chemist to the Highland and Ag. Society of Scotland, had occa- 
 sion to investigate two soils which had become " clover-sick," 
 and he caused them, together with similar adjacent soils which 
 still produced clover, to be most minutely analyzed. Without 
 reproducing his figures, which may be found in the Trans, of the 
 Highland and Ag. Soc. for 1849-51, p. 204, we will merely 
 quote some of the remarks which accompany the analyses : " The 
 results of these analyses are certainly of an unexpected charac- 
 ter, and appear to me to indicate that, in this instance the failure 
 of the clover cannot have been dependent upon the chemical 
 constitution of the soil. In both cases the results of the analyses 
 of each pair do not present a greater difference than would be 
 obtained from the analyses of two portions of soil from different 
 parts of any field." 
 
 In the present year, Stoeckhardt (Chemischer AcJcersmann, 
 No. 2, 1861, p. 85), "has published an account of several " clover- 
 sick " soils from Schlanstaedt, which reveal to analysis a greater 
 content of every nutritive mineral ingredient both soluble in water 
 and in acids, than exists in another soil from Frankenstein which 
 produces clover and wheat as well. What proves beyond a 
 doubt that the inabihty of these soils to yield clover depends 
 upon something besides their chemical constitution, is the fact 
 that lucerne and esparsette still flourish upon them admirably, 
 and further, clover itself, if sown with one of these last men- 
 tioned crops, succeeds very well. 
 
 A great truth in agriculture is this: Each kind of agricultural 
 plant requires that its seeds be surrounded with certain conditions 
 in order that they may germinate readily and healthfully, so 
 
[249] Geological Surveys of Kentucky and Arkansas, 17 
 
 that when the mother cotyledons are exhausted, the young plants 
 shall attack the stores of food in the soil with that vigor which 
 is needful in order to appropriate them without hindrance. 
 
 The fact that winter wheat is more delicate and fastidious in 
 its infancy than most other crops, is perhaps the main reason 
 why it does not succeed well on many good lands, and why it 
 cannot be continuously produced from the same soil, year after 
 year. It is a matter of experience that wheat requires a rather 
 firm seed-bed : beans, oats and mangold-wurzel approach wheat 
 in their requirements, while barley, peas and turnips are best 
 suited in a light tilth. On the other hand, climate, weather and 
 tillage so influence the character of the soil, that even on light 
 lands, wheat may find all the conditions of its growth. The bed 
 which is produced by inverting a clover sod, and allowing it to 
 consolidate by time and rains, or by passing a heavy roller over 
 it, is eminently adapted to wheat, even on a rather light soil. 
 
 The fact that in the cases given above from Stoeckhardt, clover 
 succeeded when sown with lucerne or esparsette, would indicate 
 that, possibly, the condition of the seed-bed was the cause of 
 failure. 
 
 These and other facts which might be adduced to almost any 
 extent, indicate sufl&ciently that chemical analysis alone, even if 
 we admit its full nicety and accuracy, can at the best furnish us 
 with a knowledge of but a few of many conditions which must 
 cooperate in profitable agricultural production, and as a conse- 
 quence, its part in guiding the farmer is but very subordinate. 
 Taking into the account its evident uncertainty and clumsiness 
 when applied to estimating the minute quantities which affect 
 vegetable growth, the part it can play becomes still more subor- 
 dinate — we hesitate not to say, insignificant. 
 
 As we write, a fragment from a Scientific Journal brings to 
 our notice a discovery which if real, strengthens our views in 
 an unexpected manner. It is well known that iodine is so im- 
 mensely diluted in sea-water — the soil of marine-plants — that 
 none of our tests though they are among the most delicate, serve 
 to detect it directly, and it is doubtful if it has been detected 
 even in the highly concentrated mother liquors which remain 
 after separating the crystallizable salts, yet the fuci find and 
 accumulate it, and we must grant that it is present there for 
 them, in sufficient quantity. 
 
 Again, Prince Salm Horstmar several years since, in his ad- 
 mirable researches on the influence of the individual mineral 
 ingredients of plants on the development of oats and barley, 
 found that he could not by any possibility exclude chlorine from 
 his experimental plants. His soils and pots, the salts and water 
 he fed his plants with were so purified that he could not detect 
 this element in them, and yet he invariably discovered it in the 
 
 Ail. Jour. Scr.— Second Seeies, Vol. XXXII, No. 95.— Sept., 1861. 
 32 
 
18 S. W. Jufinson on the Soil-analyses of the [250] 
 
 ashes of the plants. So too he found titanic acid in the produce 
 grown on the most carefully purified soils. Now, it is mentioned 
 in the " Chemical News " that he finds a few hundredths of lithium 
 are indispensable to the ripening of barley. This element Bunsen 
 has but recently shown to be everywhere distributed, yet it has 
 been hitherto entirely unnoticed in all soil- and plant-analyses 
 because of its occurrence in almost infinitesimal quantity. 
 
 It must be well borne in mind that Agriculture herself — so- 
 called Practice — is able of her own resources to judge somewhat 
 of the value of soils, is able to know if a soil be fertile or poor, 
 is able to pronounce upon its adaptation to crops, and can to a 
 certain extent decide what is a good manure for this or that field. 
 
 We are free to assert that the knowledge which is now to be 
 gathered from experience, is able in ninety-nine cases out of one 
 hundred, to give a more truthful verdict as to the capacity of a 
 soil, that any amount of analysis, chemical, mechanical or other- 
 wise, can do. We would give more for the opinion of an old 
 intelligent farmer than for that of the most skilled chemist in 
 most questions connected with farming. Doubtless the farmer 
 would make some blunders from which chemistry might save 
 him, but the chemist would be likely to do more violence to 
 agriculture, than the farmer would to chemistry. 
 
 By these statements which may, but should not surprise some 
 of our scientific friends, we merely intend to express an opinion 
 as to the present relative position towards agriculture of those 
 who regard the art from a chemical, and those who see it from an 
 experimental point of view. 
 
 If any one has fuller and more inspiring notions of the import- 
 ance of science in its applications to agriculture than we have, 
 we desire to sit at his feet and share the higher afflatus. But our 
 inspiration, if it be of the sort that works enduring benefit, must 
 be based on clear ideas of the directions in which advance is 
 possible and on a full perception of the difficulties that lie before 
 us, and the means of overcoming them. 
 
 We have great faith that chemistry and that chemical analysis 
 have done and are to do a work for agriculture, that shall lay that 
 venerable art under everlasting obligations to the youthful sci- 
 ence. But not by soil-analyses alone or mainly is this to be 
 achieved. We do not assert that soil-analysis is worthless — we 
 believe that the probabilities of its uselessness in direct applica- 
 tion to practice are so great that we would rarely base any ope- 
 rations on it alone, and yet it may in many cases, promote science 
 and give us data for conclusions that are of practical use. But 
 for these purposes it must form part of a system of observations 
 and trials, must be a step in some research, must stand not as the 
 index to a barren fact, but as the revelator of fruitful ideas. 
 
 We hold that soil-analysis long ago pla3''ed out the part which 
 Dr. Peter would have it perform. In the hands of Sprengel it 
 
[251] Geological Surveys of Kentucky and Arkansas. 19 
 
 was fertile with new truth, but it must henceforth be a tool for 
 occasional use, and not an engine of discovery. With our ad- 
 vance in knowledge there must be an advance in methods of 
 finding out the unknown. Soil-analysis was indeed a means of 
 insight into the secrets of vegetable growth, but it carried with 
 it the measure of its limit. What we call telescopes do not ena- 
 ble us to see the end ! 
 
 To study the soil in the hope of benefitting agriculture, we 
 must regard all its relations to the plant. We must examine it 
 not merely from those points of view which theoretical chemistry 
 suggests, but especially from those which a knowledge of prac- 
 tical agriculture furnishes. This is becoming more and more 
 the habit of agricultural chemists and the results are of the 
 happiest kind. 
 
 Let us remember what the illustrious Nestor of Agricultural 
 Science, Boussingault, has said as the summing up of his pro- 
 tracted experience and study. 
 
 " At an epoch not far distant it was believed that a strict 
 connexion existed between the composition and the quality of 
 arable soil. Numerous analyses shortly modified this opinion as 
 too positive. The sagacious Schiibler even sought to prove in a 
 research that has become classic, that the fertility of a soil de- 
 pends more upon its physical properties, its state of aggregation, 
 power of absorption, &c., than upon its chemical constitution." 
 
 " The physical properties, in my opinion, do not enable us, 
 more than the chemical composition, to pronounce upon the de- 
 gree of fertility of the soil. To decide this point with some meas- 
 ure of certainty, it is indispensable to have recourse to direct ob- 
 servation ; it is necessary to cultivate a plant in the soil, and 
 ascertain with what vigor it developes there: the analysis of the 
 plant afterward intervenes usefully, to indicate the kind and quan- 
 tity of the elements that have been assimilated." — ("i^e la Terre 
 Vegetale consideree dans ses EffHs sur la Vegetation^'''' page 283 of 
 '•^Agronomie, Chiniie agricole et Pkysiologie.^ Tome 2^remier, I860.") 
 
 There has been much progress made in our knowledge of the 
 soil during the last ten years. This advance has not consisted 
 in reveaHng to us the presence of new elements (lithia perhaps 
 excepted), nor in fixing with any more certainty the quantitative 
 limits which separate barrenness from fertility, it has not shown 
 what is the composition of a Silurian or a Sub-Carboniferous, a 
 Drift or a Tertiary soil, it has not defined the soil adapted to 
 wheat or that productive of clover, it has not indicated the ma- 
 nures which this or that soil needs; but content with the fact 
 that all soils which naturally support vegetation contain the 
 elements of vegetation, it has sought to ascertain in what forms 
 these elements are assimilable, how they may be made available, 
 what changes or reactions in the soil affect its productiveness; 
 how fertilizers act indirectly (their influence often having no 
 
20 5, W. Johnson on Soil-analyses, etc. [252] 
 
 relation to any supposable direct action), how the soil affects the 
 life of the plant otherwise than by feeding it, &c. &c. 
 
 We are approaching in fact by slow degrees to an understand- 
 ing of the physiological significance of the soil, a grand result 
 to which chemistry and physics cooperate. 
 
 We trust that in the future, the American people will not less 
 but more appreciate the value of science in its practical and es- 
 pecially its agricultural bearings; that here, as in Germany, 
 France and England, the labors of those who seek to unite Prac- 
 tice with Science may be fostered and sustained. But to this 
 end scientific men must be cautious that in endeavoring to help, 
 however honestly and laboriously they may work, they do not 
 hinder. 
 
 Sheffield Scientific School, August 20th, 1861, 
 
LIBRARY OF CONGRESS