472 BOTANY OF THE LIVING PLANT 



on the other hand, appear to arise from qualities ni some way 

 impressed upon the gametes, and so transmitted to the offspring. 

 The line between these two types of variation is often sharply drawn, 

 and it has been stated dogmatically that characters acquired in the 

 lifetime of the parent are never transmitted to the offspring. 

 Without going so far as that, it is permissible to state that hitherto the 

 evidence of the inheritance of characters acquired during the lifeti?ne of 

 the parent is insufficient. On the other hand it is believed that 

 mutations which can be transmitted from parent to offspring have 

 played a great part in Evolution. They provide a basis upon which 

 Natural Selection can work. 



Various " adaptations " of structure to external conditions have been 

 described in Chapter X. Instances have also been cited of parallel develop- 

 ment, or homoplasy, where in series of plants distinct by Descent, similar 

 adaptive modifications are seen (pp. 177, 193,210, etc.). Such results are usually 

 set down to the selection of favourable divergences from type out of inheritable 

 variations, or mutations, produced at random. But the prevalence of parallel 

 development, and even of convergence of characters, suggests that the muta- 

 tions upon which they have been built up were not produced at random. 

 That possibly the origin of " mutations " may be directed by some internal 

 or ph3'siological necessity. Tliat they may ha\-e been promoted or actually 

 determined in their direction, or their number, or their quahty, in some way 

 by the external conditions. In fact that they may have been acquired, — 

 either in the course of the individual life, or cumulatively of a succession of 

 hves, — as consequences of the impress of external conditions. The question 

 of the ori.gin of mutations, great or small, is still a quite open one. It is better 

 to entertain the possibility of some such acquisition of mutations as a working 

 hypothesis than positively to deny it. This is one reason wh}' the doctrine of 

 inheritance of acquired characters has not been ruled out in the preceding- 

 paragraphs. Its negation by Weismann was based chiefly on zoological 

 evidence. The early segregation of the germ-cells in the animal body weighed 

 greatly with him. But it needs to be stated that in Plants early segrega- 

 tion does not occur. In them the tissues, still undifferentiated as somatic 

 and germ-cells, are for long exposed to whatever the conditions of hfe may 

 be before the gametes are speciahsed. This suggests that Plants would be 

 particularly favourable subjects for observation in testing this question. 



IVIendelian Segregation. 



It had long been known that offspring produced by the crossing 

 of closely related forms, whether of Animals or Plants, does not always 

 come true to type. But it remained for Mendel to discover the 

 laws, since verified by many observers, which operate in the distri- 

 bution of the characters of the parent forms among the offspring. 

 The lollowing description of one sample of Mendel's experiments is 



