228 CHARLES W. METZ 



these figures (100, 101) is the evident polarization of the 

 chromatic threads. This appears to be characteristic of very 

 early prophases, although such stages are seldom clear enough 

 to draw. When this polarity is compared with that shown by 

 telophases (figures 111 and 112), it is difficult to avoid the con- 

 clusion that the two are correlated, — that is, that the chromo- 

 somes reappear during prophase in the same relative position, 

 and polarized in the same manner as in telophase. Prophases 

 in other species similar to those cited above are represented by 

 figures 70, 71, and 72 (Musca), 106-108 (Fucellia), 78-80 (Phor- 

 mia), 91, 92 (Sarcophaga), 130, 131 (Anthrax lateralis) and 123 

 (Eristalis). These are all essentially alike and involve corre- 

 sponding chromosome groups. Prophases, together with meta- 

 phases for comparison, in species having fewer chromosomes, are 

 shown in figures 14 and 15; 11 and 4, 5; 34, 35, 36, and 4, 5; 

 and 165 and 166. 



As seen in the figures all stages subsequent to the condensation 

 of the chromatic elements in early prophase are easily followed, 

 although the behavior of the chromosomes differs slightly in 

 different cases. Usually the association of the two members 

 of a pair becomes loose long before contraction is completed. 

 At this time the two threads are loosely and irregularly coiled 

 about one another (figs. 34, 59, 71, 92), and as contraction pro- 

 ceeds they become more and more loosely associated (figs: 35, 

 72, 93, 94). Occasionally, however, a close association is re- 

 tained up to a very late period of contraction (figs. 36, 66, 108, 

 etc.), with the consequent production of figures which very 

 closely simulate those of haploid groups. Such figures as these 

 might readily create the impression of haploid groups in diploid 

 nuclei. By the time spindle formation takes place the chromo- 

 somes are usually distinctly disjoined from their mates, although 

 the paired association is still conspicuous and may be very close 

 (figs. 15, 41, 68, 117, 132, etc.) Occasionally the process of 

 separation has been carried on so far that pairing is very in- 

 definite (figs. 153, 161), but such cases are decidedly exceptional. 

 Soon after the chromosomes become arranged on the spindle 

 they begin to show evidences of splitting in preparation for 



