MULTIPLE CHROMOSOMES 575 



species of H. viridis would not be fulty met by the explanation 

 through segregation and recombination, without assuming a 

 heterogeneous character of the material or a lack of survival of 

 certain classes. While both of these possibiUties exist they are 

 remote, and the phenomena, taken in their entirety, are much 

 more suggestive of a reconstitution of relations in the new indi- 

 vidual, probably at its origin on fertilization. This is a prob- 

 lem which concerns the most fundamental activities of the 

 chromatin substances, and about which we lack information on 

 all the stages of fertilization and maturation of the egg. Breed- 

 ing experiments should serve to decide between the possibihties 

 suggested, and these will be carried out, but the nature of the 

 processes will not be clear until the complete cycle of the chromo- 

 some complex has been studied. This fact has been realized 

 since the beginning of the w^ork on the Orthoptera, but the 

 technical difficulties involved are great and so far have not been 

 overcome. 



WTiile, therefore, it is not possible to draw^ any definite conclu- 

 sions from our present knowledge, it may be desirable to direct 

 attention to the conditions which confront us. Reasons are 

 given elsew^here for considering the indi\dduality of the chromo- 

 somes as w^ell established by the observed uniformity of num- 

 ber, size, form, behavior, etc., but in the case of the multiple 

 chromosomes we face a series of variations. As has been 

 pointed out, variability is a universal property of living matter 

 and is a subject for investigation, not an evidence of lost indi- 

 viduality. Taking into consideration, first, the accessory chro- 

 mosome, whose individuality is marked so definitely as to be 

 unquestionable, w^e find, that, within the species H. viridis, it 

 may possibly vary slightly in size although this may be only 

 apparent. From this we would draw the conclusion that, for 

 the species, size is a fairly safe criterion of homology, although 

 not absolutely exact. Using size to identify homologues, then, 

 we discover that the accessory chromosome is joined to the 

 largest chromosome in class 2 (plate II), to the third or fourth 

 from the largest in class 3, and to the fifth from the largest in 

 class 4. With persistent union, segregation and chance com- 



