1905.] NATURAL SCIENCES OF PHILADELPHIA. 



169 



globules, as roughly indicated in fig. 29; this last phenomenon I have 

 found to be of general occurrence in spermatocytes, but it has been 

 investigated most in ovocytes. The chromosomes through these stages 

 shorten and condense, some into bent or straight rods, others into more 

 or less closed rings; the surface of the chromosomes remains rough 

 and somewhat filamentous until the nuclear membrane disappears 

 (compare fig. 30 with fig. 31). The longitudinal split of the chromo- 

 somes gradually narrows, as one sees in the series of figures 23-25 and 

 27-30; it does not widen out, so that the relations of tlie univalent 

 components of a bivalent chromosome remain approximately the same 

 as in preceding stages. The early bivalent U or V of the synapsis 

 period may become a straight dumbbell, or its univalent arms may be- 

 come apposed along their length, or it may become a ring; but in all 

 cases the position of the longitudinal split is along the length of each 

 univalent chromosome, whether that be straight or bent (figs. 27, 28, 

 30). Very rarely have the chromosomes an X-shape (fig. 31). There- 

 fore each bivalent chromosome is composed of two univalent chromo- 

 somes joined by one end or by both ends (in the case of rings), and the 

 space enclosed by a ring is not a longitudinal split but the area sepa- 

 rating two entire univalent chromosomes. Wliere the two univalent 

 chromosomes of a pair are connected is in most cases marked by a 

 constriction (x, figs. 27, 29), and in the rings there may be two such 

 constrictions (the larger ring of fig. 29) in accordance with the con- 

 junction in these cases of both ends. These chromosomes are thus 

 essentially, in formation and shape, like those of the Hemiptera and 

 Peripatus: each represents two longitudinally split univalent chro- 

 mosomes joined by one or both ends. And the gradual narrowing or 

 closure of the longitudinal split is as evident and undeniable as in any 

 other object studied by me. 



Through these prophases the heterochromosome is recognizable by 

 its smooth contour and compact structure (Plate IX, n. 2, figs. 27-30). 

 It is now almost always in the form of a rod so bent that both arms lie 

 contiguous and parallel, as shown in fig. 26 where the arm seen on high 

 focus is stippled and that seen on deeper focus drawn in outline only. 

 Each of its arms, as we have seen, represents a univalent heterochr^ - 

 mosome. 



With the disappearance of the nuclear membrane, which commences 

 to dissolve^ away first at the poles near the centrosomes (Plate IX, fig. 

 31), the chromosomes have attained their completed dense structure 

 and smooth outline and take their position within the equatorial plate 

 (figs. 32-34) . There are exactly ten bivalent chromosomes present, one- 



