The Spermatogenesis of the Opossum (Didelphys virginiana) etc. 
Ot 
to synapsis, sensu stricto, the various workers take the opposing 
positions: that it is an end-to-end, and a side-by-side union of chromo- 
somes respectively. 
Discussion. The abundance of perfectly preserved cells with 
nuclei in which the chromatin is in synizesis leaves no question, I 
believe, as to the normal character of this phenomenon in spermato- 
genesis. The accumulation of much conforming evidence respecting 
synizesis by many workers in gametogenesis, and especially the testimony 
of Sargant (1897), Overton (1905) and Wilson (1909) that similar 
appearances present themselves in living cells, invalidates the contention 
of a number of cytologists notably, Mc Clung (1902), Schaffner (1907), 
Guignard (1899), Janssens (1905), Miyake (1905), Davis (1908), and 
Duesberg (1908) that it is an artefact. Moreover, much of the evidence 
presented by the last set of men may simply indicate that synizesis is 
not an absolutely constant process in gametogenesis. Synizesis need 
not necessarilv be constant in order that it may be normal (i. e. ; not 
a fixation artefact) when it does appear 6 ). 
It seems altogether probable that both telosynapsis and parasynapsis 
occur in different forms. As yet, there appears no evidence of a strict 
limitation of one or the other to certain groups of animals or plants. The 
Schreiner’s (1906) believe that parasynapsis is the general rule in the 
reduction process. Goldschmidt (1908) holds the opposite view, and 
points out the fact that a di-spireme occours also in gametogenesis of 
parthenogenetic forms and in certain somatic cells in early prophase, 
wliere there is no question of a synapsis. Büchner (1910) also argues 
cogently for telosynapsis as cliaracteristic of the orthoptera. 
The character of the early prophase chromosomes (double threads; 
later, loops and paired threads) suggests that perhaps no sharp distinction 
really exists between the end-to-end and side-by-side method of con- 
jugation. Here the one appears to follow the other. The reduction takes 
place first by end-to-end conjugation. Later on, the limbs of the loops 
approximate and fuse more or less completely. In this condition the 
bivalent chromosome is divided in metakinesis. Where parasynapsis 
has been described, an end-to-end fusion must still occur at the ends of 
the threads. Thus in either event the side-by-side union (primarv or 
secondary) is perhaps the ultimate rule in reduction; and the di-spireme 
condition of the postsynaptic thread and the prophase thread of somatic 
mitosis may have no other elements in common. Moreover, the fusion 
of the conjugants in synapsis may be so complete, as apparently to vitiate 
all individuality (at least visible) as the work of Bonnevie (1907 and 
