146 M. \V. BLACKMAX. 



are bivalent there is one element which is univalent in character, 

 being derived directly from a single spermatogonial chromosome. 

 This is the accessory chromosome and in the metaphase can 

 always be distinguished by its characteristic shape, and especially 

 by the fact that it is connected by mantle fibers to only one pole 

 of the spindle. There is usually no indication of the plane of 

 division in the accessory chromosomes at this time, although in 

 the prophase a longitudinal constriction is often shown. 



The chromosomes of Scolopendra heros are seventeen in number, 

 sixteen of which are bivalent, while one is univalent. The sixteen 

 bivalent chromosomes undergo longitudinal and transverse divi- 

 sions during the prophase and at the beginning of the metaphase 

 are of several different types as regards shape. After studying a 

 large number of metaphases of both the large and small (Black- 

 man, 105) type of spermatocytes it has been found to be a rule 

 that in each cell the chromosomes of the several types bear a 

 definite and constant numerical relationship to each other. This 

 fact can best be appreciated by referring to the accompanying 

 plates. It will be seen at a glance that the cruciform tetrads 

 which have been described above as type A are in all cases six in 

 number. Furthermore, among the chromosomes of this typical 

 form a definite size relation exists which makes it possible to 

 arrange the cross-shaped chromosomes in a graded series on the 

 basis of bulk. To be sure, the difference in size is not so striking 

 as that existing between the largest and smallest chromosomes in 

 some insect material, but I believe is as great as the difference in 

 size between adjacent chromosomes of the graded series in insects. 

 It is perhaps unnecessary to explain that the actual size difference 

 is in many cases greater than appears in the camera lucida draw- 

 ings, owing to the fact that some of the chromosomes are fore- 

 shortened on account of the angle from which they are viewed. 

 For this reason, the drawings are much less convincing than the 

 preparations. 



The shape of the chromosomes, aside from showing an apparent 

 modification due to the angle of vision, actually docs vary con- 

 siderably but each of the group of six chromosomes in question 

 are always reducible- to the cruciform tetrad described as type A. 

 The variations in shape have to do only with the degree to which 



