Oll tlie Dimegalous Sperm and Clu'omosomal Variation of Euschistus, etc. 137 
way fluctuation of number, but Variation with regard to mode of conju- 
gation of chromosomes. Quite similar cascs of such Variation have been 
describcd by me before (1901, 1906) but in miich less detail, for Harmo- 
stes reflexulus, Corizus alternatas, Chariesterus aniennnator, Tingis clavata 
and Oncopeltus fasciaius] and by Wilson i) for Lygaeus turcicus. 
2 . The other kind of Variation foimd is due to the occasional presence 
of supernumerary chromosomes, never niore than 2 in number, found 
only in testis no. 120. There may be in the spermatogonia, first and 
second spermatocytes either 1 or 2 of these. These supernumeraries are 
adecpial in volume, constantly a little larger than the smaller diplosome, 
exhibit therefore constant diagnostic qualities, and do not aiipear to di- 
vide in the maturation mitoses. This kind of Variation does produce an 
actual Variation in the total number of the univalent chromosomes, for 
these may vary in number from 14 (the norm) to 15 or 16. This Variation 
is due to the addition of new individuals to the normal series. The pro- 
babUity that they do not divide at the fh'st mitosis of maturation would 
indicate that they may be nnequally distributed from cell to cell, wliich 
again would account for their ajiparent absence in some cells. Further, 
this kind of Variation would result in unequal distribution of chromo- 
somes in fertdization, and in this way the supernumeraries might be pas- 
sed over to some individuals and not to others. 
3. In a single case out of 672 the 2 diplosomes failed to conjugate 
in the second maturation spindle. 
4. The phenomena of chromosomal Variation in Euschistus do not 
in any way indicate that chromosomes are variable clu'omatin complexes, 
in the sense of being reforniations of the chromatin from one cell division 
to another. On the contrary constant sizes and fornis of the particular 
univalent chromosomes reappear from one mitosis to another, oiie 
of the strongest proofs that they are persisting individualities. The 
chromosomal fluctuation is in no way chance Variation. Were the chi’o- 
mosomes temporary variable combinations of the chromatin, subject to 
rearrangements of their substance between one mitosis and another, no 
explanation would be available for the consistency of the phenomena in 
Euschistus. 
5. Such variations in the beha\’ior of germ cell chromosomes might 
well be the basis of congenital variations. And a form such as Euschistus, 
that exhibits chromosomal Variation in a fahly large percentage of cases, 
might well be in a period of species formation. The variations in the 
1) 1905. Studies on Chromosomes. I. Journ. Exper. Zool. 2. 
