THE ACCESSORY CHROMOSOME. 65 



of the element so carefully as he does in a later one ('96). Here, 

 even after more extended investigation, he is uncertain "both as 

 to its origin and its fate" but inclines to the belief that it is 

 "nucleolar substance" despite the fact that it does become 

 ''homogeneously mingled" with the chromatin. As will be 

 noted (1 23), he was at first inclined to consider it related in 

 some way to the centrosome, but as a result of more careful 

 study decided that this was a mistake. 



In addition to noting the obvious staining reaction and pe- 

 ripheral position of the element in the prophase of the sperma- 

 tocyte, Paulmier also observed its behavior in the spermatogonial 

 divisions (1 36) and subsequently in the different phases of the 

 spermatocytes (^ 37). As a result of a recognition of its very 

 apparent chromosomic character he agrees with me in calling it 

 a chromosome, but prefers to speak of it as the " small chro- 

 mosome." I have already pointed out in a previous paper ('oo) 1 

 my reasons for regarding this as a misnomer on account of its 

 usually being larger than the other chromosomes so need not 

 again refer to the subject. Thus, we have as a result of the latest 

 and apparently most accurate work upon the Hemiptera, the his- 

 tory of a small chromosome that puts it in almost complete 

 agreement with the behavior of the accessory chromosome in the 

 Orthoptera. 



From the preceding statements of different investigators, it will 

 be apparent, I think, that there is in the spermatocytes of all in- 

 sects so far studied an unusual nuclear element which is charac- 

 terized (i) by a remarkable uniformity in staining power, similar 

 to that exhibited by chromosomes in the metaphase ; (2) by a 

 continuous peripheral position during the spireme stage, at least; 

 (3) by an isolation from the chromatin reticulum and nonpartici- 

 pation in its changes ; and (4) by fission during metakinesis after 

 the manner of chromosomes. In addition to these features which 

 have generally been recognized, there are others which material 

 of exceptionally good character has rendered apparent to several 

 observers. Of these I wish to speak later but desire here only 



1 11 3> P- ^5 ; latter part of r , 3, p. 86 ; and *| 5, p. 89 ; were attached to the proof 

 as footnotes but were included by the printer in the body of the article. It thus 

 happens that on page 89 a reference to Paulmier' s last paper ('99) appears before a 

 criticism of his earlier one ('98). 



