236 REPRODUCTION 



numbers of characters. Thus, Canna indica has six chromo- 

 somes while LiUum martagon has twenty-four; but there is no 

 evidence that HUes have four times as many characters as have 

 cannas. A still better confirmation of our assumption is found 

 in the Droseras, where of two species closely allied one has 

 twenty and the other has forty chromosomes. We can, there- 

 fore, confidently proceed with the conviction that each chromo- 

 some bears an indefinite group of characters. 



Can we with the highest powers of the microscope discover 

 organized units composing a chromosome, each of which might 

 be supposed to bear a single character? The utmost we can 

 see under the microscope of the finer structure of the chromo- 

 somes is best observable in the early prophases of nuclear divi- 

 sion. Here with the most favorable subjects it can be seen in 

 preparations properly stained that the nuclear thread is made 

 up of deeply stained granular masses or chromatin, alternating 

 with unstained portions called linen. The chromatin masses, 

 as has been said, have been termed ids (Fig. 129, i in A and B). 

 It has been proposed that an id is composed of various single 

 character bearers or pangenes and that an id therefore rep- 

 resents a group of characters. With this partly theoretical, 

 partly observational ground to build on a very plausible hy- 

 pothesis has been erected to explain the occurrence of all pos- 

 sible combinations of characters in the offspring of hybrids, 

 and the possibility of fixing these combinations so that they 

 will come true to seed. Briefly stated it is this: In a hybrid, 

 during the prophases of nuclear division of microspore and 

 megaspore grandmother cells it is possible that an interchange 

 of pangenes may take place between homologous paternal and 

 maternal chromosomes, one chromosome being considered ho- 

 mologous to another when its pangenes are so matched with 

 the pangenes of the other as to make pairs of contrasting 

 characters. This hypothesis is in harmony with the results of 

 hybridization, and it does not conflict with what can be seen 

 under the microscope. 



Theory of Pangeneic Interchange. — Let us turn to the 



