lojJ cmtOMOSOMES IN THE SPERMATOGENESIS OF THE HEMIPTEKA HETEROPTERA. 



,u:;ivo us the means of recognizing, after the determination of tlie chromosomal pairs, tlie 

 maternal and paternal chromosomes of each nucleus, and thereby advance our means 

 of analysis still another step. 



>\.nd a word may be added here to those who may be sceptical as to the possi- 

 bility of distinguishing particular chromosome pairs. Any one who looks over the 

 plates given in this paper, and notes the chromosome pairs distinguished by corre- 

 sponding letters, may say that the imagination plays too large a part in such distinc- 

 tions. But he should recall that we can draw no conclusions without the help of the 

 imagination, and that what we see we must also imagine. But more than this, he 

 should i-ecall that the printed figui'e can in no way be as clear as the preparation under 

 the focussing microscope since it can reproduce only the profile, whereas the eye sees 

 this and also the depth of the structure. One has only to draw the chromosomes care- 

 full}^ with the camera lucida, then search for correspondent ones upon such drawings, 

 to be convinced of the actual presence of such pairs. And above all, no one has any 

 right to express doubt of these relations who has not made broad comparative obser- 

 vations of his own. 



This constant difference of the chromosome pairs, and the probable con.stant 

 th(jugh much slighter diffbrences of the elements of each pair, which are the exj^ression 

 of both murphological and physiological distinction, I would denote by the term 

 '■ chromosome difference " which expresses the phenomena jierhaps a little more pre- 

 cisely than Boveri's term "nuclear constitution." 



■^>. The Number of Chromosomes and Taxonomy. 



One incentive to me to make comparative studies of the chromosomes in the 

 Hemiptera was to determine how far the number of chromosomes is constant in a 

 particular group of animals; and certain conclusions were presented in two preceding 

 papers (1901r/, 19016). From the observations on the Hemiptera then made it 

 appeared that the chromosomal numbers were not constant, so that tlie determination 

 of the factors governing the number seemed as unexplained as ever before. And in 

 now touching on the question again 1 find that the problems are as difficult of solu- 

 tion as ever. 



Yet it seemed wortli while to reexamine the matter from a taxonomic standpoint, 

 to test the value of chromosome numbers as criteria of racial affinity. And since no 

 one has tal)ulated the number of chromosomes known in animal species, not since the 

 brief list of cases summarized by Wilson (1900, pp. 206, 207), I have compiled these 

 statistics for the germ cells only of the greater immber of described species ; there are 

 a number of omissions because some of the literature was inaccessible, but the list is 



