204 BOTANICAL GAZETTE [march 



but longitudinal division could proceed in the normal way and the 

 identity of the chromosomes not be lost. But in reduction the number 

 should be definite, if the karyokinesis is to furnish normal cells. 



The nucleolus is still present when the chromosomes are fully 

 developed, but often shows signs of fragmentation, as in the examples 

 shown in fig. jia. After the chromosomes are developed, the cyto- 

 plasm also shows a change in structure, having passed from a spongy or 

 reticulate arrangement to a more or less radiate structure (figs. 31, 32). 



Individuality of the chromosomes. — The chromosomes continue to 

 become more indefinite in shape until they appear as irregular, dark- 

 staining, apparently structureless masses, very unequal in size. The 

 real character of the chromosomes can be studied to advantage only 

 during the formative period, although the larger ones can be recog- 

 nized even in the mother star. In the incipient chromosome loops 

 individuality is very marked. As stated, there are three small ring 



(figs 



'1 



very prom 



(figs. 20b, 22,27, 



of various shapes and sizes. Since these are bivalent chromosomes, 



it is evident that, on the theory of the conjugation of maternal and 



must 



in shape and activity. In the microsporocytes the bivalent chromo- 

 somes have an individual shape and size easily distinguishable, and 

 the inference from this is evident, as also in the massing of the chro- 

 matin in the early prophase, that these bodies are individualized an 

 retain their individuality from one division to another. Were t e 

 chromosomes not individualized, they could not preserve such definite 

 forms and numbers from generation to generation. Dunng 

 ontogeny, the chromosome passes through a series of forms, on y 

 return, as any other organism, to a definite type at a definite stage. 

 From the present study and the investigations of others, it is evi** 

 that the mechanics of chromosome reduction is simple, the usw 

 spirem orienting itself into folds, twists, or simple loops, giving n - 

 to all the various shapes, as rings, rods, coils, tetrads, and cross*. 

 The actual form observed in any individual case may be a mere pr 

 jection, and great care should be taken to ascertain the actual s ape 

 by observation from various points of view. 



