160 CHKOMOSOMES IN THE SPERMATOGENESIS OF THE HEMII'TERA HETEROPTERA. 



gonic numbers into even ovogonic ones) may be arranged in their order of frequency 

 as follows : 



24 chromosomes is the unreduced number in 30 species, about one-sixth of the 

 whole list; the numbers 32 and 14 occur each in 24 species; the number 16 in 20 

 species ; the numbers 12 and 22 each in 9 species; the numbers 18 and 20 each in 7 

 species ; the numbers 4, 8, 30 each in 6 species ; the numbers 28 and 36 each in 5 

 species ; the numbers 10, 34, 48 each in 4 species ; the numbers 26, 40, 52 each 

 in 2 species; and the numbers 2, 38, 42, 46, 50, 60, 64, 116, 168, each in only one 

 species. 



Thus the full number of chromosomes is below 34 in the greater number of species 

 so far studied. 



Certain of these animals show the rare peculiarity of having two normal numbers, 

 one twice that of the other ; thus Ascaris megalocephala has either 2 or 4, Ascaris Iwin- 

 bricoides, 24 or 48, Helix pomatia, 24 or 48, and Echinus microtuberdatus, either 18 or 

 36. In each of these species we might distinguish then a variety " univalens " from one 

 "bivalens," as O. Hertwig (1890) has done for Ascaris megalocephala. In the last 

 form Meyer (1895) was able to distinguish no anatomical differences between the two 

 varieties, and Herla (1893) has proven that there is frequently crossing between them. 

 But such hybrids contain three chromosomes, not twice the lower normal number. 

 And evidently variation in the normal number, such as that of the four species men- 

 tioned, cannot liave originated by polyspermy, for three spermatozoa would have to 

 fertihze an ovum to produce double the usual normal number of chromosomes ; and 

 Boveri (1902) has shown that such polyspermy results in abnormal development. 



Further, two cases are known where the spermatid has a different number of 

 chromosomes from that of the ovotid, Planaria and Slyelopsis, these being cases not due 

 to the presence of a monosome in the spermatogenesis. 



Finally, let us examine the constancy of the chromosomal numbers within certain 

 circumscribed groups of animals. In some a certain constancy is to be found : the 

 normal number is 24 in all the urodelous Amphibia ; McClung (1905) states there are 

 always 23 in the spermatogenesis of the Acrididse among the Orthoptera (but Syrbula 

 and CaJoptenus are exceptions to this) ; among the Pentatomidte (Hemiptera) either 14 

 or 16 is the number (17 species examined), but Banasa has probably about 28 ; in the 

 Coreidse the numbers are 22 or 14 (one with 16) ; in all the opisthobranch molluscs 

 examined it is 32 ; and in the Turbellaria, 12, 16 (most usual), 18 or 20. In most of 

 the other groups of equivalent scope the variation in number is so great that there 

 seems to be no constancy ; thus in the hemipteran family Lygteidse there may be 24, 

 14, 16 or 28. And in the spermatogenesis of two closely related species of Gryllus 

 Baumgartner (1904) finds the numbers to be 21 and 29. 



