444 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



That functional differentiation may exist quite independently 

 of morphological differentiation is equally conspicuous in the 

 frequent cases of self- and cross-sterility where perfect anatomical 

 compatibility is present. Furthermore, functional differentiation 

 is subject to wide fluctuations and is by no means necessarily 

 related to morphological differentiation. This is seen in the 

 different grades and intensities exhibited by Linum grandiflorum, 

 Primula sinensis, and Lythruni Salicaria; by the different grades 

 that exist in numerous species of Primula; and most especially by 

 different grades and intensities of functional differentiation 

 existing within a species or within a single form as judged by 

 the immediate results of pollinations and the character and 

 behavior of the respective offspring. 



With Lythrum, for example, Darwin obtained various grades of 

 fertility from illegitimate intra-form pollinations for all these 

 forms, showing that various grades of physiological incompatibility 

 exist independently of morphological differentiation. The illegiti- 

 mate offspring exhibited a wide range of variation with respect to 

 vigor, impotence, and fertility (the latter judged wholly by cross- 

 pollinations, mostly open and legitimate). Only in one set of 

 19 plants (sisters?) was provision made for exclusive intra-form 

 pollinations ('69, p. 398, 399). As these were all of the long- 

 styled form they were illegitimate offspring illegitimately polli- 

 nated. Of these only three plants are fully reported: one pro- 

 duced a large crop of capsules and was decidedly fertile; one pro- 

 duced few capsules with less seed per capsule; the third flowered 

 profusely, no impotence is mentioned, but the plant produced few 

 capsules which, however, contained so many seeds that Darwin 

 states "the average and the maximum are so remarkably high 

 that I cannot at all understand the case." No attempt was made 

 to study controlled self-fertilization on any plants, and no attempt 

 was made to make comj:)arisons of the relati\e degrees of incom- 

 patibility which the data suggest must exist in and between 

 various plants even of the same form. The studies of illegitimate 

 offspring were also extended only to the first generation. 



With Primulas, there is. it appears, a rather weak incom- 

 l)atibility in illegitimate self- or intra-form fertilizations. Dar- 

 win's most extended experiments in inbreeding a form were with 

 the long-styled form of P. veris ('77, p. 219), in which four gener- 



