STUDIES ON INSECT SPERMATOGENESIS. 397 



Table I, into which some errors have doubtless crept through 

 insufficient material, accidents of sectioning, or doubtful synonymy 

 of the generic titles. Where the material was obviously insufficient 

 or unsatisfactory the fact is indicated by a dash, and doubtful 

 points are marked " ?." It should be added that, like Montgomery, 

 I have as a rule recorded size differences only when they were reason- 

 ably obvious. In other words, a case like Murgantia in which the 

 size differences are slight, might be overlooked, especially if the 

 material was not first class. Furthermore, I have made no special 

 effort to detect lobes in which the spermatocytes are unusually 

 small. This was because the differences in this respect are by no 

 means striking, — in fact are often rather difficult to make out even in 

 Euschistus if the material is poor, — and the value of the results 

 " seemed hardly commensurate with the labor involved. 



The table gives in the first column the list of forms examined, the 

 nomenclature followed being that of Van Duzee ('17); in the second 

 column is given the number of lobes in the testis of a given species, 

 while the third and fourth columns give the index numbers of the lobes 

 in which unusually large or unusually small (where noted) cells occur, 

 these numbers being assigned in accordance with the plan previously 

 described (see Text-figs. lA and B). Comparisons were generally 

 made on spermatocytes in the late growth period, and exact observa- 

 tions on the mature sperm were usually impossible for obvious reasons. 



Examination of Table I shows at once that far from being a phe- 

 nomenon unique to a single genus, the spermatic polymegaly is widely 

 distributed in the Family Pentatomidae, and is in fact rather more 

 common than rare. Montgomery's original idea is therefore erroneous, 

 and his failure to detect differences in some of the other genera which 

 he examined for this point is rather puzzling. Thus in our common 

 species of Brochymena the difference is very striking, though Mont- 

 gomery reported a general equality for this genus. Possibly a clue 

 is to be found in the case of Nczara, in which no dimegaly occurs in 

 the northern species, hilaris, but a very noticeable one in the southern 

 species, viridula. At the same time it is interesting to note that Van 

 Duzee ('17) has now placed these two species in different genera, 

 Nezara and Acrosternum. Doubtless the form which Montgomery 

 examined was Acrosternum (= Nczara) hilaris, in which absence of 

 dimegaly is also indicated by my observations. Further, in a number 

 of known cases the identification of forms studied by Montgomery 

 ('98 and '10) was not very critical, and the discrepancies may be due 

 in part to such errors. 



