MATURATION PHENOMENA OF GERM CELLS. 147 



Now for Protenor I found that in the spermatogonic chromo- 

 somal plate there are always exactly thirteen elements (Fig. 22) ; 

 the two smallest of these are the heterochromosomes marked 

 N, n they are paired, and in the following synapsis stage conju- 

 gate to form the smallest bivalent chromosome N, n of Fig. 23 ; 

 these two chromosomes are then quite similar to the hetero- 

 chromosomes (chromatin nucleoli) of the other Hemiptera. But 

 there is a large element in the spermatogonium (X, Fig. 22), 

 unpaired there, and which does not conjugate with any other 

 chromosome during the synapsis stage, but remains unpaired in 

 the spermatocyte (X, Fig. 23). This element I called the 

 " chromosome .r." Now, as McClung has also pointed out, this 

 chromosome behaves exactly as does an accessory chromosome 

 in the Orthoptera, being unpaired in the spermatogonia, and not 

 conjugating with any other chromosome during the synapsis 

 stage. Therefore in Protenor occur both kinds of heterochromo- 

 somes, the small paired ones, A 7 and ;/, and the large unpaired 

 one, X. N and n are "chromatin nucleoli" according to my 

 terminology, while X is an "accessory chromosome"; thus 

 both kinds of heterochromosomes occur in the same cell, and 

 their likenesses and differences were fully discribed by me for this 

 species. Both are recognizable through the whole growth period 

 of the spermatocytes by their compact form and deep staining ; 

 but only the small pair, N and n, can be recognized in the rest 

 stage of the spermatogonia. In three other Hemiptera, Alydits, 

 Hannostcs and CEdancala I showed that in the spermatogonia 

 occurred also an uneven number (thirteen) of chromosomes ; but 

 in these the odd chromosome does not maintain its compact form 

 during the growth period of the spermatocyte, and so is not 

 recognizable there ; but in the first maturation mitosis it is 

 immediately recognizable as the only chromosome that has not 

 conjugated with another to form a bivalent one. Now such a 

 chromosome of A/ydits, Hannostcs and CEdancala agrees with 

 the chromosome X of Protenor in not pairing with another during 

 the synapsis stage ; but differs from it in behaving like the other 

 chromosomes after a mitosis, i. e., in losing its compact structure 

 and strong affinity for chromatin stains. 



Why should a heterochromosome be sometimes unpaired in the 



