Ill 



giving, however, the same results, viz. : the formation of tetrads 

 by a primary longitudinal and a secondary transverse division of 

 the spireme. The first method {Cyclops, CantJiocamptiis) agrees 

 almost exactly with what I have here described as the " rod type." 

 The spireme is divided longitudinally before segmentation into 

 half the normal number of chromosomes. The chromatin then 

 begins to collect into a much thicker double rod ; the rod then 

 divides transversely and the tetrad is formed by longitudinal and 

 transverse division. The other method described by Rijckert 

 takes place in Heterocopc and Diaptomns and agrees very closely 

 with what I have described as the " ring type." The double 

 spireme breaks as before, but the double segment, instead of re- 

 maining contiguous throughout, becomes separated in the middle, 

 while the ends alone remain in contact. A ring is thus formed 

 and tetrads arise later by two divisions, one through the diameter 

 of original cleavage, the other at right angles to this line. 



Enough has been given in this account to show that certain 

 methods tetrad-formation are characteristic of certain species of 

 animals. Up to the present time it has always been found that 

 the tetrads in a single nucleus are formed by one method, either 

 by the " rod type " or by the " ring type " alone. My observa- 

 tions on the fern, however, show that tetrads in the same nucleus 

 may be formed by both methods or even by a third. From these 

 facts the conclusion seems inevitable that all the types of tetrad- 

 formation mentioned above are merely modifications of the same 

 process and have no significance in themselves so long as a com- 

 mon result is obtained. 



It seems remarkable that such obvious structures as tetrads 

 should have been hitherto overlooked in the plant reproductive cells. 

 Many observers have noticed that the mitosis in the sporocyte 

 differs from that of all other cells, whether somatic or archesporial. 

 This difference was early recognized by Guignard and Strasburger. 

 Overton and Belajeff also were struck by the peculiar shape and 

 appearance of the chromosomes in this mitosis ; the latter especi- 

 ally described them as agreeing in all particulars with Flcmming's 

 account of heterotypical mitosis. Finally Farmer (1895), in speak- 

 ing of the pollen-mother-cell of Liliuni Martagon, refers to 



