My Dear Sir,

I am extremely sorry that so long a time has elapsed before I have been able to answer yours of Feb 16.¹ It happens that besides occupation for the press and other matters which could not be postponed, we have been, and are, extremely busy sampling for future analysis the soils of nearly all of the more than 20 plots on which our experiments on the effects of different manures on the mixed herbage of grassland have now been conducted through twenty seasons—of the results of which you know something.² We take to 6 depths of 9 inches each, or in all to a depth of 54 inches, from 3 places on each plot, mixing the 3 of each layer together as one sample, & in this way we have taken about 150 specimens, & hope to collect more yet.

With regard to soil-matrix, such as you require for the experiments you propose, I imagine that a rather clayey soil, first well burnt with free access of air, & then well washed several times with distilled or clean rain water, would provide sufficiently favourable physical conditions, & at the same time be poor enough in plant food to answer your purpose.³

In our paper in the Phil. Trans. part 2. 1861. pp. 470–472, you will see described our method of preparing such a matrix (p 471).⁴ At p 472 there is a description of the mineral constituents supplied and at p 542 of the Ammonia solutions added; and in subsequent experiments of the same kind we used weak solutions of Nitrates, the strength of which I will look up, if of interest to you.

Our object was of course to supply all the cinereal constituents that could be needed, and in favourable condition for the most normal growth, whilst your object is, I presume, to restrict the supply of one or another constituent, so as to force the plant to abnormal courses, so far as may be consistent with the perfection of seed? You are doubtless aware of the experiments of Stöckhardt, Sachs, Knop, Nobbe, Liegert, E. Wolff, and others, on Water-culture; of which there is a short notice in Johnson’s “How Crops grow”.⁵ Their experience may afford you some hints as to strength of solutions, &^c; and, if I remember rightly, both Maize & Oats were found to give seed, which again gave perfect plants, with no other silica supplied than that in the original seed (and a little accidentally provided), the plants themselves being of course almost destitute of it.⁶ In this fact is evidence that considerable modification can be submitted to. Then, again, in our experiments on “Mixed herbage” grown by different manures for many years in succession, we have most remarkable variation in the composition of the ash according to the artificial supply even though in a natural soil.⁷ Thus, whilst in cases in which potass is freely provided by manure we have from 35 to 40 per cent of potass in the ash of the mixed produce, with perhaps only 5 per cent, or less, of Soda, although it also is liberally provided, we have without potass in the manure in some cases as little as 15 per cent of potass in the ash, with even more Soda that potass; and what is more remarkable, in some instances of great exhaustion of potass, we have not only a large proportion of Soda, but more than 1 per cent of Lithium. In reference to these extreme variations it should be observed, however, that as in the experiments in question from 30 to 40 species are growing together, each with its own special habit, requirements, root-range, &^c, such only will be prominent as can best submit to the conditions, & they will have their tendency of growth modified from that of normal growth accordingly. Indeed, a most obvious effect of such exhaustion is a great reduction in the number of prominent species, & much reduced tendency to form flower & seed, or even flower-stem, the herbage consisting chiefly of very dark green foliage (if nitrogen be freely supplied), a character which always betokens an excessive percentage of Nitrogen, or in other words a defective assimilation of carbon. Although, therefore, chemical analysis shows such a wide range of variation in the composition of the mixed not ripe herbage, the character of the growth induced is unfavourable to the supposition that the one constituent can replace the other in a normally developed & well matured plant. In this connection I may mention that Nobbe, at Tharand,⁸ has been growing plants in water, excluding one or other base, or one or other acid, and if my memory serves me, potass, lime, & phosphoric acid, at any rate, were essential to the perfection of seed, in all the cases, & the development was more or less abnormal according to the exact nature of the exclusion. Upon the whole I am disposed to anticipate that as your object is to experiment on generation after generation, to which end not only life but health & seed formation are essential, the range of variation in the supply of the mineral constituents which will be admissible will be less than the results of our experiments above referred to might at first sight be supposed to indicate. Nevertheless, when it is borne in mind that, stated in general terms, relatively excessive Nitrogenous supply may be said to give luxuriance, that is quantity of growth, longevity, &^c, & relative excess of the mineral matters tends to maturation, that is to restrict luxuriance, & favour flowering & seeding, it is obvious that the tendencies of the plant may be much varied by varied supply.

We have had some incidental notes taken on variation on the mixed herbage plots, but I have no doubt that a careful examination of them by those who know better than we do what to look for might bring to light interesting & important facts. I need not say that both M^r. Lawes⁹ & myself would be very glad to see you or your son¹⁰ or both here for that purpose. May or June is of course the best time—will your son come sooner? Such an examination of changes actually taking place under known conditions, including those of the domination among the species, the tendencies of development, and the coincident variation in chemical composition might usefully aid the choice of means to effect the changes you seek to investigate. It is, however, one thing to have at command the varying habits of from 30 to 50 species growing together, the resultant of whose growth or failure is a close reflection of the conditions provided, affording indeed a searching analysis so to speak of the available matters within the soil, and quite another to deal with an individual species, & to succeed in providing conditions under which it will readily and healthily grow and mature, & at the same readily vary.

With regard to constituents to be supplied, I would suggest very weak solutions of Nitrate of potass, Soda, Lime or Magnesia, or Lithium. With these it would probably not be desirable to supply nitrogen in the form of phosphate of ammonia also—instead the phosphoric acid might be given in combination with the fixed base or bases it is intended to supply; but whatever phosphate you use it should not have an alkaline reaction, which could be remedied by dilute sulphuric acid. I may mention that ammonia-salts are more frequently injurious to plants than Nitrates; but if you supplied phosphate of ammonia, you might then supply potass, Soda, Lime or Magnesia, as sulphates.

I imagine that in such experiments many failures must be anticipated—that is, that although you may get active vegetation up to a certain point, the tendency of the exclusion of normal constituents, will be rather to cause abnormal development of such portions of the plant as can best bear the privation than to induce the development of perfect but varied flowers & seeds.

I may mention for what the observation may be worth, that not only Iron but Manganese has been a pretty constant constituent in some of our ashes, & that dealing with such a large number of ashes, the examination of the Manganese precipitate from many mixed together has shown determinable quantities of Nickel & Cobalt. Solutions of Iron, Manganese, copper, nickel, or cobalt, and even of of Cæsium & Rubidium might perhaps, even with otherwise normal ash-constituent supply, at any rate affect colour. I do not know what organic fluids to suggest, but some aniline compounds¹¹ might be tried.

I do not know whether Hopkins & Williams¹² devote attention to other than Pharmaceutical preparations—if not, I think you would get all you want from Mess^rs J. J. Griffin & Sons. 22 Garrick Street Covent Garden.¹³

I imagine you might burn a clay soil with some little organic matter in a heap in the open air as weeds, &^c. are burnt, & be able to remove the burnt clay free enough from the vegetable ashes, to yield, after frequent washing, a sufficiently poor matrix. The burnt clay should be put into a tub, well stirred with water, the latter drawn off, & so on frequently. If you find difficulty perhaps we could prepare you some soil at the laboratory here—indeed M^r. Lawes begs me to say that we shall be glad to render you any assistance.¹⁴

Hoping these somewhat desultory remarks may be of some service, I am My Dear Sir | Yours sincerely | J. H. Gilbert

See—Über die organische Leistung des Kalium in der Pflanze, Nobbe, Schroeder und Erdmann. Versuchs-stationen Organ, Band XIII. 321 & 401.¹⁵

Charles Darwin Esq^re, &^c. &^c.