MATURATION PHENOMENA OF GERM CELLS. 151 



that its two univalent elements pass to opposite cells, so the sper- 

 spermatid has nevermore than one of them. One of those in the 

 spermatogonium must accordingly have come from the spermatid 

 and one from the ovotid, which combined to give rise to that 

 spermatogonium. And here, also, in the formation of such a 

 bivalent chromosome in the synapsis there must be a union of a 

 paternal with a maternal chromosome. No other explanation 

 seems possible. 



These conclusions, the numerical ratios of certain clearly dis- 

 tinguishable chromosomes in the spermatogonia, to certain 

 equally distinguishable ones in the spermatocytes and sperma- 

 tids, could be established for the heterochromosomes for some 

 forty species of Hemiptera, and for other chromosomes in the 

 cases of TricJipcpla scinivittata, Protcnor bclfragci, Pcliopclta ab- 

 brcviata, Prionidus cristatns, Zaitlia fluminea and Corizus latera- 

 lis. To make this point clear a few figures of certain of these 

 cases are given here again. In the spermatogonium of Anasa 

 (Fig. 16), as well as in the ovogonium (Fig. 15), are recognizable 

 among the 22 chromosomes, two very much smaller than the 

 others (A", ;/, heterochromosomes), and two considerably larger 

 (A, <?); in the first spermatocyte there are eleven bivalent ones, six 

 of which are shown in Fig. 17, and here we recognize again the 

 chromosomes N, n and A, a. In Corizns, in the spermatogo- 

 nium (Fig. 1 8) are two particularly small (A 7 , ;/) and two partic- 

 ularly large chromosomes ; and these recognizable again in the 

 first maturation spindle (Fig. 19), Similarly in the case of Tricli- 

 pcpla (Figs. 20, 21). In the spermatogonium of Protcnor (Fig. 

 22) are thirteen chromosomes ; a particularly large one (A'), two 

 next in size (K, k\ and two smallest (N, 11). In the spermatocyte 

 (Fig. 23) A" is recognizable as being the largest; it is the odd 

 chromosome and is not paired with any other. A^ and // are 

 paired and so are K and k. 



From these observations I concluded that probably in every 

 case the chromosomes in the synapsis united to form bivalent 

 pairs in such a way that the one of each pair was paternal and 

 the other maternal ; and I was able actually to demonstrate it in 

 those cases where the differences in volume between the chromo- 

 somes were sufficient to allow them to be followed from genera- 



