1905.] NATURAL SCIENCES OF PHILADELPHIA. 199 



cases of bastardization between different species, as described by Guyer 



(1902) and Moenkhaus (1904), the maternal and paternal chromosomes 

 fail to conjugate. Or, if the parents have different numbers of chromo- 

 somes, some of those of the parent with the larger species are forced to 

 remain univalent during the growth period, as shown by Rosenberg 



(1903) for Drosera. Facts like these might suggest that the presence of 

 heterochromosomes has been produced by bastardization of species 

 with different number of chromosomes. But that could be the case 

 only of unpaired heterochromosomes; it would not explain the paired 

 ones, and we have found that the unpaired kind are probably derivable 

 from the paired. Again, they have been found in all insects in which 

 they have been sought for, or in nearly all, but it would be rash to con- 

 clude that all these species of insects have arisen as bastards between 

 parental forms with different chromosomal numbers. Therefore there 

 is no good reason to refer the heterochromosomes to any hybridization 

 process; and every reason to consider them as modified conditions of 

 the ordinary chromosomes, formed in some cases concomitantly with a 

 change in chromosomal number, probably from a higher number to a 

 lower, chromosomes with a different metabolic activity and on the way 

 to disappearance. A remarkable fact, for which I see no explanation 

 whatsoever, is their very general occurrence among insects, and their 

 absence elsewhere except in spiders; but they may be found in other 

 groups when the attention is given them that they deserve. McClung 

 (1902a) has put out the hypothesis that they are sex-determinants, 

 reasoning from the condition of the unpaired heterochromosome of Xiph- 

 idium; here only half of the spermatids receive the division products, 

 and he argues that its presence in them may determine the male sex. 

 This is only a hypothesis, and as yet we do not even know whether in 

 the ovocytes of such species similar heterochromosomes may not occur. 

 Indeed, whether spermatozoa with and those without heterochromo- 

 somes are equally capable of fertilization is not known, and would be 

 exceedingly difficult to determine. Further, in some species of Hemip- 

 tera all the spermatozoa receive division products of the heterochromo- 

 somes, and on McClung's hypothesis all spermatozoa in such species 

 would produce males. 



On the question of the perpetuation from generation to generation 

 of an odd number of heterochromosomes or ordinary chromosomes I 

 have touched at another place (1901&); but now I am convinced it is 

 inutile to discuss this problem until we have facts of their behavior in 

 the maternal germ cells. 



In conclusion, attention should be drawn to the recent divergent 



