388 Gates and Rees . — A Cytological Study of 
Batrachoseps and Agar’s (1911, PI. II, Figs. 12, 13) on Lepidosiren on the 
other, it seems probable that the difference in opinion regarding these forms 
is entirely one of interpretation of this stage. We might be inclined on 
this basis to conclude for the telosynaptic account in all, but for the fact 
(1) that the telophase and prophase split in the spireme stage of the somatic 
chromosomes is by no means universal, and (2) that Hogben (1920), whose 
figures of these stages in Periplaneta are too diagrammatic to be of great 
value as evidence, nevertheless has the distinct merit of having counted the 
loops in the bouquet stage in oogenesis and found the diploid number of 
loops before the pairing and the haploid number afterwards. This appears 
to show definitely that parasynapsis occurs in Periplaneta. 
The Orthoptera among animals were long supposed to show telo- 
synapsis,but the latest account in these insects (Wenrich, 1916) gives strong 
evidence in favour of parasynapsis in Phrynotettix. Owing to differences 
in staining and in form, it was possible to trace one pair of chromosomes, 
and to a less extent two other pairs, into the late semi-resting telophases 
of spermatogonial divisions and from the earliest heterotypic prophases 
through the spireme stages to diakinesis. This method of tracing the 
history of a single chromosome might be supposed to yield absolutely 
crucial evidence, but we believe the evidence just falls short of this. 
Wenrich’s Figs. 68, 69 (PI. VI) show the pair of more deeply staining ‘A’ 
chromosomes in the early heterotypic prophase, and like all the prophase 
chromosomes in Phrynotettix he describes them arising, not by the approxi- 
mation of lateral halves, but as an irregular spiral in a looser matrix. 
His following Fig. 70 is interpreted as showing the conjugation of the 
pair of ‘ A ’ chromosomes. But this ‘ pair ’, which consists of two parallel 
beaded threads, is apparently thinner than either of the single spiral ‘ A ’ 
chromosomes of the preceding stage. Important and valuable as this 
paper is, we cannot therefore regard it as a demonstration of parasynapsis, 
although the evidence as a whole certainly favours that interpretation. 
The nature of synizesis and the occurrence of binucleate pollen mother- 
cells have already been sufficiently discussed in the text. 
Chromosome Twisting and Chiasmatypy . 
The cytological phenomenon described by Janssens (1909) as chias- 
matypy has been much discussed in recent years as the possible physical 
basis of the genetical phenomena of crossing-over, particularly in Drosophila 
(Morgan, 1919). As Wilson (1920) has recently pointed out, the original 
scheme of Janssens involved four strands in a bivalent chromosome, and 
Morgan (1919) has shown how on the parasynaptic theory with four 
parallel strands various distributions of the segments of the four chromatids 
might result. Wilson, who adheres to the parasynaptic method of heterotype 
chromosome formation in animals, concludes that Janssens s interpretation of 
