300 BOTANICAL GAZETTE [APRIL 
through the various stages in advance of others, a condition com- 
parable to that noted by JANSSENS (21) in Batracoseps. The 
appearance at later stages seems to bear out this supposition. Fig. 
46 shows a split in a portion of the spirem, while the rest appears 
entirely undivided. Fig. 47 shows several places where the parts 
have become still more widely separated, yet a considerable mass 
of the nuclear material is still in synapsis; and fig. 49 shows all 
gradations from paired loops to typical gemini. 
This irregularity in development, together with the fact that 
there is no definite succession of stages in the loculi, makes a fully 
satisfactory interpretation of the spirem stages impossible. [If it 
is true, however, as seems probable, that the chromatic loops which 
appear as the mass emerges from synapsis consist of a single thread 
which afterward shortens and thickens, and if the split shown in 
figs. 46 and 47 represents a separation of threads previously paired, 
the series accords closely with GrEGOIRE’s hétérohoméotypique 
scheme. Figs. 44 and 45 would then represent the pachyténe stage, 
the loops shown in fig. 45 resulting from a shortening and _ thicken- 
ing of thinner loops shown in fig. 44. Fig. 46 doubtless represents 
a splitting of the pachyténe loops, the diploténe stage. Figs. 47 
and 48 show different stages in strepsinema; in fig. 47 only a small 
part of the nuclear mass is in the strepsiténe condition, while fig. 
48 is a more advanced stage. 
The appearance of fig. 43, which is quite characteristic of the 
loosening of the synaptic knot, makes it appear possible that the 
spirem comes out of synapsis as a series of paired threads which 
afterward fuse to form a continuous spirem. That this actually 
occurs in most cases is held by SrrasBuRGER and his school. 
Such a condition as is shown in fig. 44, however, where the loops 
appear still thin but with the halves widely separated, makes it 
seem probable that the thickened spirem arises from the thinner 
by a thickening of the threads. That is, figs. 44 and 45 represent 
merely different phases of the pachyténe stage. 
3In referring to this stage as characterized by the occurrence of loops consisting 
of a single thread, the writer does not mean to enter into the discussion as to whether 
this ‘‘single thread” is ay * unit or is composed of two separate threads twisted 
together. The expression “single thread” is used to mean simply a loop which 
appears as one thread, as distinguished from one composed of two parallel threads. 
