THE SPERMATOGENESIS OF LEPIDOSIREX PAEADOXA. 23 
transverse division — have now been shown not to have the 
significance that has so long been attached to them. The 
recent works of Lerat, and especially of Matschek, have 
demonstrated that the transverse joint in the tetrads has no 
relation to the division plane of either division, the chromo- 
somes splitting longitudinally both times. Thus in the early 
anaphase II the separating daughter-chromosomes are trans- 
versely jointed like those of anaphase I. The joint dis- 
appears in the late anaphase. This matter is returned to on 
p. 29. 
Intercalation of a Resting Stage in the Meiotic 
Pro phase. 
The series of stages just described undoubtedly represents 
the normal course of events. Sometimes, however, the stages 
figured in figs. 18-23 are not immediately followed by 
the disappearance of the nuclear membrane and re-pairing 
of the chromosomes, but instead, the nucleus enters into a 
sort of resting stage, passing through the condition shown in 
PI. 5, fig. 34 to that figured in fig. 35. The chromo- 
somes, univalent and in the full somatic number, take up their 
positions equidistant from one another under the nuclear 
membrane, and fine threads appear joining them up. For a 
time they retain the appearance typical of the diakinetic 
chromosomes, but presently this shape is lost, and they 
become irregular plates of chromatin connected up with one 
another by numerous bridges. In most cases thirty-eight 
entirely separate chromosomes can be counted with ease, but 
often some of them show a more or less pronounced paired 
arrangement. 
These nuclei are extremely striking objects, especially when 
seen with the stereoscopic eyepiece, owing to the distribution 
of the chromosome as thin plates close under the membrane, 
leaving the interior of the nucleus entirely free from any 
staining substance. A faintly staining nucleolus is present. 
Although on looking through a number of sections a good 
