CHROMOSOMES IN THE SPEiaUTOGENESIS OF THE HEMIPTEUA HETEUOrTEIiA. 159 



For purposes of comparison the chromosomal numbers of the spermatogonia (and 

 ovogonia), or those of the ovotids (and spermatids), are the safest to consider, because 

 in cells of these generations in almost all cases the chromosomes are univalent, while 

 different observers have varied greatly in their estimates of the valence of chromosomes 

 of the ovocytes and spermatocytes. It is probable that the spermatogonic (or ovogonic) 

 number of chi'omosomes is always double that of the number in the spermatid (or 

 ovotid), so that the one can be readily calculated from the other ; the only exception 

 is in cases of spermatogenesis with a monosome, where the spermatid may contain one 

 more chromosome or one less than half the number in the spermatogonium. And 

 for purposes of comparison the full (not reduced) number of chromosomes is prefer- 

 able, because in any species all the spermatogonia have the same number of chromo- 

 somes, while the spermatids may have different numbers. 



Wilson's discovery that when there is an uneven siDcrmatogonic number of 

 chromosomes in the spermatogenesis there is an even number in the ovogenesis intro- 

 duces a complexity in the comparisons. But this is easily obviated ; for so far as 

 known when the spermatogenesis lias an uneven number it contains always one 

 chromosome less than the ovogenesis, therefore, e. g., a spermatogonium having 13 

 chromosomes we can calculate the ovogonium to have 14. In such cases we will use 

 for comparison only the number of the ovogenesis, whether directly ascertained or 

 whether derived by adding one to the spermatogonic number when the latter is an 

 odd one. 



When we look over the statistics presented in these tables we find that the num- 

 ber of chromosomes of the ovogonium or spermatogonium (translating odd spermato- 



