20 BOTANICAL GAZETTE [jaxtary 



{iigs. i8j ig) ; these two stages may 



5 



INI\\ 



The abo\'e brings the miter to the conclusion that the spirem is 

 of double nature from the outset, probably derived from the paternal 

 and maternal source, as was interpreted by Allen (i, 2, 3), Gre- 



GoiRE (25), Bergh (6, 7, 8, 9), Rosenberg (37,38), and cIrdiff 

 (12). 



In preparations stained with iron-aIum-hematox\din and extracted 

 so as to obtain a proper differentiation, the pah-ed chromatin threads 

 of the spirem appear as if consisting of darkly stained granular parts 

 and a Kghtly colored ground-substance of threadlike structure, but 

 this simply meajis that they are not of homogeneous thickness : the 



more agsre 



.... jated the chromatin substance, the more dark appears 

 the stain, and therefore probably the individualit}- of the two halves 

 of the spirem may not be lost entirely durbg the dose association. 



: TTitre are observed in Nephrodium a number of parts of the spirem 

 nmning through the contracted mass from the nuclear membrane 

 where the mass lies in contact {jig. 18). Gregoire (24) found that a 

 sumlar condition of the spirem exists in Osmunda, and that in a later 

 stage the chromosomes resulting from the thick spirem held a similar 

 orientation in regard to the nuclear membrane. From this he con- 

 cluded that there is no continuous thick spirem in the prophase of 

 heterotypic division of Osmunda. Since in Nephrodium the thick 

 spirem as it emerges from synapsis is always in contact with the nuclear 

 membrane up to.the segmentation of chromosomes, the condition 

 shown m fig. 18, where the number of parts of the spirem run out 

 from one side of the membrane, may have some significance which 

 we are unable to interpret at present. 



The thick spirem emerging from s>Tiapsis showed first, at the 



_ widely uncoiled from the contracted mass, longitudinal fis- 



sions which came undoubtedly from the separation of the two inde- 

 pendent threads that entered into association during s)-napsis. The 

 thick spirem simply segments into 64 or 66 bivalent chromosomes, 

 each of the pairs being derived from the independent thread. In a 

 certain stage before. the "strepsinema". condition, appearances of 



ooping were also observed in connection with the contracted 



they had nothing to do with the formation of chromosomes 



