MULTIPLE CHROMOSOMES 565 



genus and related genera. In a recent paper Robertson ('16) has 

 considered this matter with great care and has advanced strong 

 arguments in support of the position that the family number 

 may be maintained in the face of apparent reduction. His argu- 

 ment may be thus summarized. (1) The number of chromo- 

 somes for the family is twenty-three in the male; (2) the form 

 of the chromosomes, when uncombined, is that of a rod with 

 terminal fiber attachment; (3) size relations are constant and 

 valid indications of homologies; (4) in the presence of less than 

 the typical number of chromosomes certain ones of the com- 

 plex are Vs with non-terminal fiber attachment; (5) if the limbs 

 of the Vs are counted as units, the number twenty-three is re- 

 stored ; (6) the behavior of the limbs of the Vs in maturation is 

 parallel to that of corresponding free elements in other species 

 of the family; (7) the morphological composition of the V is 

 indicated by a non-chromatic bridge at the angle where the fiber 

 attaches. 



I beUeve that in the species studied by Robertson the criteria 

 are largely applicable and the conclusions essentially justified. 

 The evidence is, to a considerable extent parallel to that of H. 

 viridis, mth the exception of intermediate steps in linkage. 



There is clearly a permanent reduction in the number of free 

 chromosomes — it then becomes a question as to whether there 

 is a definite loss of morphological entities in the chromosome 

 complex, or whether all the twenty-three elements are poten- 

 tially present, even though in combinations. It must be recog- 

 nized, I think, that, however fundamentally alike are the cases 

 of Hesperotettix and Chorthippus, they stand on a somewhat 

 different basis morphologically. In the former, independence 

 of certain chromosomes is strongly evidenced by their free con- 

 dition in some individuals and their partial union in others; in 

 the latter the number of free elements is constant, not only for 

 this genus but for others nearly related, and e\ddence of com- 

 binations must be more indirect. Considering the criteria which 

 I have employed in Hesperotettix and Mermiria I would regard 

 the following as probably applying in the case of Chorthippus: 

 (1) The number of chromosomes, so strongly indicated in the 



