1905.] NATURAL SCIENCES OF PHILADELPHIA. 197 



bivalent in the spermatogonium — there as a chromosome pair with the 

 components closely united instead of being, as with most of the chromo- 

 somes, separated. Can we go further than this, and consider the un- 

 paired heterochromosomes of the Orthoptera to be also already bivalent 

 in the spermatogonium, but with the univalent parts so closely united 

 that they do not become separated even in the reduction mitosis? The 

 heterochromosomes of the Orthoptera appear to be usually larger than 

 the ordinary chromosomes, which is the only observational evidence 

 for the idea that they may have the value of more than one chromo- 

 some, and sometimes they are much larger. Such evidence is, of 

 course, not at all sufficient. But should they be ultimately proven to 

 be bivalent in the spermatogonia, a further uniformity would evince 

 itself: all heterochromosomes and all ordinary chromosomes would be 

 paired in the spermatogonia, whether the two members of a pair be 

 separated there (umValent) or be united (bivalent) ; in the former case 

 they would become bivalent by conjugation for the first time in the 

 spermatocytes, in the second case they would pass over already bivalent 

 to the spermatocytes. In any event an even number of univalent 

 chromosomes in the spermatogonia and half that number of bivalent 

 ones in the spermatocytes would be the primitive (unmodified) condi- 

 tion, as it is the one most usually found. In the preceding paragraph 

 it was shown to be probable that the odd ordinary chromosomes of the 

 Hemiptera and the unpaired heterochromosome of Protenor are already 

 bivalent in the spermatogonia; this may or may not be the case with 

 the unpaired heterochromosomes of the Orthoptera, but if it is the 

 case, as I think is somewhat probable, then the following conclusion is 

 reached — a conclusion well based at least for the odd ordinary chromo- 

 somes and the unpaired heterochromosome of Protenor: heterochromo- 

 somes that are paired in the spermatogonia and become bivalent in the 

 spermatocytes would be an earlier condition, and would lead to the 

 later condition of heterochromosomes unpaired in the spermatogonia 

 by conjugation of their univalent components in spermatogonic cell 

 generations. In this way unpaired heterochromosomes would be later 

 modifications of the paired ; and in the same manner, unpaired ordinary 

 chromosomes later modifications of paired ordinary chromosomes. 

 Two univalent chromosomes of a spermatogonium might conjugate to 

 form one bivalent one before the spermatocyte stage, this would then 

 be an odd ordinary chromosome, which later might or might not become 

 an unpaired heterochromosome; or two ordinary chromosomes of a 

 spermatogonium might become heterochromosomes (chromatin nucle- 

 oli) but still remain univalent in this cell (conjugating not before the 



