166 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



chrysanthemums &c. ; such might be called "gains," but they are accom- 

 panied by "losses" of the essential organs. The law of compensation 

 seems thus to interfere somewhat with De Vries' definitions. 



It is true De Vries adds, ' No great changes have been attained 

 without acquiring new qualities on one side and reducing others to 

 latency" (p. 222). But he does not seem to allow for the fact that 

 "latency" itself may disappear. Thus, the water crowfoot has acquired 

 the dissected type of submerged leaf, like almost all dicotyledonous 

 aquatic plants, by the influence of the water. When, however, the seeds 

 are grown on land, though the anatomy resumes the ancestral aerial 

 characters, the external morphology is retained, and is therefore constant 

 and specific. The "latent " completed leaf of land forms has gone. 



He alludes to " the conception of latency of characters as the common 

 source of the origination of varieties " (p. 242) ; but this is not Darwin's 

 view, as he seems to imply, for Darwin attributes them to the influence 

 of " changed conditions of life," leading to " definite or indefinite results," 

 i.e. the variations in the offspring. 



In raising his mutants, elementary species and varieties, De Vries 

 observes : " Seed-cultures are henceforth to be considered as the sole means 

 of recognising the really existing systematic results of nature" (p. 61) ; 

 but he overlooks the fact that culture with a new and prepared soil is 

 the very best means of breaking up Nature's constant species, as we 

 shall see in his own cases. 



Let us consider his examples. CEnothera Lamarckiana is a garden 

 plant, long cultivated, and unknown, at least to De Vries, as a wild 

 species, only occurring in Europe as an escape, if not cultivated. It is 

 a plant requiring insect fertilisation, though rare instances of self- 

 fertilisation have been seen. It may then, perhaps, have been a hybrid. 

 When he discovered it in a field in 1886, both stems and rosettes (being 

 a biennial) were seen to be highly variable, and soon distinct varieties 

 could be distinguished among them, but many were too weak to live 

 a sufficiently long time in the field. De Vries transplanted the rosettes 

 into his garden, with the well-known results. 



The impression left on one's mind is that several of his " forms," to 

 use an indefinite expression, had nothing more exceptional than what 

 may happen with any cultivated plant, which has broken down its natural 

 stability of character, and that the cause was the various conditions of 

 the soil in which it had been previously, and was by himself, cultivated. 



The two "strong species," as he calls them, gigas and rubrinervis, 

 remained constant from their first appearance ; and he observes : " Con- 

 stancy is not the result of selection or of improvements ; it is a quality of 

 its own " (p. 562). 



But it has been shown above that, to secure constancy, the form must 

 be grown in the same conditions which produced the variation. One 

 asks, therefore, was not gigas grown continuously under the same con- 

 ditions ? If so, its constancy might have been foreseen with considerable 

 probability. 



Gigas. — Stem stouter but not taller than that of (E. Lam. : inter- 

 nodes shorter ; leaves more numerous ; flowers closer together ; corolla 

 larger ; fruit only one-half the size ; seeds fewer but larger. 



