462 THE AMERICAN NATURALIST [Vol. XLIV 



than to dislocation of organs as his view in the strictest 

 sense seems to require. It should not be forgotten that 

 eggs normally fertilized if kept under unfavorable con- 

 ditions so that they develop abnormally show similar im- 

 perfections. Were his results really due to dislocations, 

 i. e., mal-assortments of chromosomes, we should antici- 

 pate a far greater mosaic type of development, I think, 

 than actually appears. 



In conclusion we must consider the behavior of the 

 chromosomes at that period in their existence that has 

 seemed to most cytologists the most critical time in their 

 history, especially in relation to their behavior in hered- 

 ity. I refer to the so-called synapsis, when the total 

 number of chromosomes becomes reduced to one half the 

 number characteristic of the body-cells. The most sig- 

 nificant fact in this reduction is that like-chromosomes 

 pair, or unite, as first made probable by Montgomery, 

 and since confirmed on an extensive scale by several 

 other writers, notably by McClung, Wilson, Stevens, 

 Schreiner, etc. 



It may appear that we can most easily interpret this 

 process as due to like materials running together or 

 fusing in consequence of the likeness of the materials 

 themselves. "But that the process is something more 

 than this seems probable from the fact that such union 

 takes place at no other time in the innumerable resting 

 stages, except at this particular one, just prior to polar- 

 body formation in the egg, and at the corresponding 

 period in the spermatogenesis. The actual apposition 

 of the thread-like chromosomes that has been described 

 by many observers does not suggest a simple physical 

 fusion or running into a lump of like materials, but 

 rather the approach and fusion of definite cell constit- 

 uents. The line of separation persists for some time in 

 some species, according to certain observers, and may, 

 according to Brauer and others, remain evident until the 

 next division occurs, when the threads again separate to 

 pass to different parts of the spindle. 



The mechanism appears to be such, on this interpre- 



