ASSOCIATION OF CHROMOSOMES IN DIPTERA ' 225 



show the chromosomes clearly) . Above them is a diagram show- 

 ing the two groups in position as they appear in the section. 

 Each chromosome in the one group is seen to be represented 

 by a corresponding sister chromosome similarly oriented in 

 the other. Such cases furnish unequivocal evidence that the 

 two members of a pair are not daughter halves of a univalent 

 prophase element, but are distinct chromosomes, and that they 

 both divide equationally in metaphase. 



In the third place, diploid groups in the males of species hav- 

 ing an unequal X-Y pair, demonstrate by the morphological 

 difference between X and Y that the pair is composed of two 

 distinct chromosomes. A striking example of this is seen in 

 the three species of Drosophila shown in figures 41, 42, 44 and 

 45 (compare with figs. 49 and 50) in which species the X-chromo- 

 some of the males is fully twice the size of its mate Y. It would 

 be difficult indeed to imagine these being daughter halves of a 

 univalent chromosome. The same features are also brought 

 out by other species having unequal sex-chromosomes (figs. 85, 

 86, 88, 124, 135, 137, etc.), although the evidence is not always 

 so striking as in the three species cited. 



These lines of evidence, I believe, leave no escape from the 

 conclusion that pairing of chromosomes is a reality in the species 

 here considered. That the mosquitoes are no exception to this 

 rule will be shown below when the different groups of flies are 

 treated independently. 



The essential difference between the above results and those 

 of Taylor and of Lomen center around one particular feature — 

 the behavior of the chromosomes in late metaphase and early 

 anaphase. The other stages are not seriously disputed. The 

 question, therefore, is whether the two metaphase elements 

 separate from one another in anaphase, thus effecting a reduc- 

 tion division, as described by Taylor and by Lomen, or whether 

 each divides and sends a daughter half to either pole as Stevens 

 maintained. I believe that I have demonstrated the correct- 

 ness of the latter conclusion in the above paragraphs, and need 

 not dwell further on it. The difficulty in -the work of Taylor 

 and Lomen is due, I believe, to faulty fixation of their material. 



