508 Edmund B. Wilson. 



has endeavored to bring together under the name of "hetero- 

 chromosomes" two classes of chromosomes in these insects, 

 namely, the "unpaired heterochromosome" ("accessory chro- 

 mosome" of McClung) 1 and the "paired heterochromosomes" 

 (or "chromatin nucleoli"), which differ markedly in behavior 

 from the other chromosomes during the maturation process. 

 Montgomery gives as the most essential characteristic of these 

 chromosomes "their difference in behavior from the other chro- 

 mosomes in the growth period of the spermatocytes and ovocytes, 

 as sometimes during the rest period of the spermatogonia, a dif- 

 ference which appears usually to consist in the maintenance of 

 their compact structure and deep-staining intensity, so that while 

 the other chromosomes become long loops or even compose a 

 reticulum, these do not undergo any such changes or only to 

 slight extent" ('05, p. 191). "Thanks to this peculiarity they 

 can be followed with extreme certainty from generation to genera- 

 tion, even during rest stages; and so are splendid evidence for the 

 thesis of the individuality of the chromosomes" ('04, p. 146). 



The study of these chromosomes has led Montgomery to some 

 very important conclusions regarding synapsis and reduction with 

 which, as far as their more general features are concerned, I am 

 glad to find my own results in substantial agreement. Considered 

 more in detail, however, there are many points regarding which 

 I think Montgomery's general treatment of the "heterochro- 

 mosomes" requires emendation. 



In a preceding paper (Wilson, '05) the fact was indicated that two 

 types of "paired heterochromosomes" or "chromatin nucleoli" 

 occur in Hemiptera. The first, including what I have called the 



'Since there is no reason for considering the "accessory chromosome" as in any sense accessory to 

 the others, it appears to me that McClung's term might well be abandoned in favor of a less com- 

 promising one. I suggest that until their physiological significance is positively determined chro- 

 mosomes of this type may provisionally be called hete r ot'op;c chromosomes (in allusion to the fact that 

 they pass to one pole only of the spindle in one of the maturation-divisions) in contradistinction to 

 amphitropic chromosomes, the products of which pass to both poles in both divisions. There are several 

 objections to this term, one of which is that the "accessory'' chromosome behaves as a heterotropic 

 body in only one of the divisions (and probably in one sex only). Another is the fact that the members 

 ("chromatids") of every chromosome-pair are heterotropic in the reducing division, since this only 

 separates univalent chromosomes that were previously in synapsis; but if, as in these studies, the term 

 "chromosome'' be consistently applied to each coherent chromatin-element of the equatorial plate, 

 whatever be its valence or mode of origin, this objection is perhaps not serious enough to weigh against 

 the convenience of the term. 



