McAllister—Cytology and Embryology . 615 
“angular spirem stage” of Eisen. The tension on the nuclear 
membrane causes the nuclear cavity to grow smaller at this 
stage, but finally the spirem becomes broken apart at the weak¬ 
est places which are the points of th© connection of the chromo¬ 
somes. Janssens is of the opinion that this stage is improperly 
called second synapsis. 
Van Molle (57) in an investigation of the spermatogenesis 
of the squirrel divides synapsis into three stages, “lepto-synap- 
tine,” “amphi-synaptine” and “pachy-sy nap tine.” His Figure 
13, representing a thick split spirem massed in the center of the 
nucleus, is interpreted by him as the latter stage, but is. in every 
way a good representation of the second contraction figure. 
Arnold (3) in his Figures 12, 13, and 14 of the spermatogen¬ 
esis of Hydrophilus piceus represents a stage which he inter¬ 
prets as post-synapsis. The chromatin, according: to Arnold, 
is aggregated in the center of the nucleus in a dense knot, while 
from this loops radiate into th© surrounding clear space in a 
manner characteristic of the second contraction. Arnold how¬ 
ever calls these loops “linin” though they are figured as strands 
as thick as spirem strands. 
Schleip (88) in the ovogenesis of Hotodromas monarcha 
shows in his Figures 66 and 67 a stage following the uniformly 
distributed spirem after synapsis, in which the major part of 
the spirem is aggregated in the center of the nucleus with radi¬ 
ating loops of a split spirem. It is a typical second contrac¬ 
tion figure and coincides with this stage as to time as well. 
Grates (30), referring to the fact of the non-occurrence of the 
“bouquet stage” in the higher plants, is of the opinion that “the 
second contraction phase probably corresponds to it.” 
I have reviewed the literature pertaining to the so-called 
atpical methods of embryo-sac formation in an earlier paper 
(50). Since the appearance of this paper Miss Stephens* pre¬ 
liminary note (93) on the Penaeaceae has been followed by a 
more detailed account (94). She has found that the 16-nucle^ 
ate embrvo-sac of this group arises by four divisions of the em¬ 
bryo-sac mother cell,—her earlier observation that an axial 
