150 CHROMOSOMES IN THE SPERMATOGENESIS OF THE HEMIPTERA HETEROPTERA. 



dissimilarities of form and size in the unmodified chromosomes, tlie autosomes. I 

 showed (\Q01b) that in a number of species of Hemiptera there are spermatogonia 

 chromosome pairs marked by pecuHarities in size ; and that when this is the case there 

 are corresponding bivalent elements in the first spermatocytes, i. e., that these size dif- 

 ferences are constant during succeeding cell generations. I also showed in the same 

 memoir that chromosomes of like size conjugate in the synapsis stage, and proved that 

 of the two chromosomes that so conjugate the one is paternal and the other maternal, 

 consequently that the synapsis is to be interpreted as the last stage in fertilization, the 

 conjugation of correspondent chromosomes of opposite nativity. In the next year 

 Sutton (1902) showed that in Brachystola all the autosomes compose pairs. And then 

 (1904a) I demonstrated that in the spermatogonia of Urodelous Amphibia the twenty- 

 four autosomes can be without difficulty resolved into twelve pairs, the components of a 

 pair being distinguishable not only by size relations but also by peculiarities in form ; 

 and I showed this to be true of Ascaris also, where the ovotid contains one small and 

 one large chromosome and the spermatozoon introduces one small and one large one. 

 Wilson (1905) has recently found this to be the case for a number of Hemiptera, adding 

 materially to ni}' former observations ; and in the present paper this constancy of pairs 

 in the spermatogenesis is detailed for a still greater number of species. We can say 

 that whenever the chromosomes are not too small or too numerous, they can be seen 

 to present certain size relations that remain constant during succeeding cell genera- 

 tions, united sometimes with certain form relations as Baumgartner (1904) also has 

 shown. McClung has likewise found this to liold true for certain of the Orthoptera. 



So we are justified in saying that each spermatogonium and ovogonium has a 

 double series of chromosomes, a paternal and a maternal set, which go to make up a 

 series of pairs, the pairs being of gradated sizes or forms, and each pair composed of a 

 paternal and maternal element of approximately equal size and form. The two ele- 

 ments of a pair probably lie close together in tbe spirem stage of the spermatogonium 

 as I showed elsewhere (1904a) ; and even in the equatorial plate they frequently lie 

 close together. Tlie two elements of such a pair are the ones that conjugate in the 

 synapsis stage, and tbat separate from each other in the first maturation division. 



Accordingly, even where there are no such great differences present as between 

 autosomes and allosomes, distinct pairs can frequently be distinguished, and thereby 

 morphological differences of size and form be made out. It is obvious that chromo- 

 somes of different sizes cannot have the same phj^siological value ; they must have 

 activities difl^ering at least in amount. But we may decide tbat their activities differ 

 also in kind, else; a particular chrorno.some would not alwa3's conjugate only with its 

 correspondent in form and size but should be expected to conjugate with any other 



