STUDIES ON CHROMOSOMES 71 



usually outside the ring (as in fig. 2h) ; while the third ra-chromosome 

 is always coupled with the two others at the center of the ring, and 

 moves to one pole without division. In the second division these 

 relations are almost exactly reversed, the ra-chromosome dividing 

 equationally as a univalent, while the supernumerary does not 

 divide and is typically coupled with the idiochromosome bivalent 

 near the center of the group. I desire to emphasize the fact that 

 these differences are in no way obscure or difficult to see, but are 

 conspicuously shown in so great a number of cells as to remove 

 all doubt. 7 



7 This point demands emphasis because of the scepticism expressed by certain 

 writers in regard to the constancy of the chromosomes in respect to number, size 

 and behavior. Conspicuous among these writers is Delia Valle ('09) who has 

 brought together a valuable if somewhat uncritical review of the literature, and 

 contributes careful observations of his own upon variations in the chromosome- 

 number in the somatic cells of Salamandra. Such scepticism is perhaps not sur- 

 prising in view of the unlucky contradictions that still exist in the literature even 

 of so favorable and well known a group as the insects. But to ascribe this con- 

 fusion of the literature to a confusion of the facts i. e., to an inconstancy so great 

 as to preclude the possibility of attaining exact results would be, I think, a fatal 

 blunder. The confusion in the literature cannot, of course, be attributed altogether 

 to mistakes of observation or to accidents of technique though both these must 

 be held to a strict reckoning. I am not aware that anyone has maintained that 

 the relations of the chromosomes form an exception to all other biological phenom- 

 ena in being absolutely fixed and immutable; and due weight should be given 

 to the numerical variations that have been recorded by Delia Valle and many others 

 myself included. The fact remains that it is possible to determine accurately 

 what are the normal or typical relations of the chromosomes, as of other struc- 

 tures, and to establish in many cases their high degree of constancy. The same 

 common sense must be used in the treatment of these relations as in the case of 

 other phenomena that are subject to variation. For example, insects have been 

 seen with seven legs, but it is not for this reason to be doubted that insects have 

 six legs. In like manner, in the ovaries of Largus cinctus I have seen as many as 

 three dividing cells that show 13 chromosomes; but I nevertheless do not doubt, 

 after the study of a large number of cases, that the typical number is 12. 



The case of Metapodius is disposed of by Delia Valle in the following easy fash- 

 ion. "Not constancy but variability in the number of chromosomes is the general 

 rule in all organisms; of which the observations published by him (Wilson) are 

 but a special confirmatory case" (op. cit., p. 161, translation). Better acquaintance 

 with the facts in Metapodius would probably render Prof. Delia Valle less certain 

 of this; for I am confident that no observer of ordinary competence could confuse 

 such a series of relations as that here displayed with the occasional fluctuations 

 with which we are familiar in many forms, including this very genus. 



