6 ARTHUR BOLLES LEE 
and find that neither these nor any other of the animal chro- 
mosomes that I have studied do so; but that on the contrary, 
at one period of their existence, they all do possess a cer- 
tain spiral differentiation answering, to some extent, to 
Vejdovsky’s ‘chromonema’. The following pages set forth 
the evidence for this, but will, as I think, also show that the 
advocates of the chromonema theory have pushed it too far ; 
for the spiral differentiation in question does not constitute 
an independent fibre, and does not form the germ of a new 
chromosome. 
The chromosomes described are chiefly those of prophases, 
equatorial phases, anaphases, and telophases; but I have 
touched on those of some interphases in which certain of their 
characters are demonstrable. I do not attempt in this paper to 
describe the nuclein elements of completely * resting’ nuclei. 
The results set forth are based on the study of chromosomes of 
the Amphibia (chiefly Urodela). Careful investigation of the 
nuclei of the other classes of the Vertebrata has shown that their 
chromosomes, though conforming apparently in all respects with 
those of the Amphibia, are mostly too small to afford trust- 
worthy images of the details in question. The same is the case 
with most of the Invertebrata, only certain nuclei of the Ortho- 
ptera beimg found to possess chromosomes which, though 
smaller than most of those of the Amphibia, yet afford images 
which are often clearer. The majority of the figures are of 
chromosomes of spermatogonia, the most favourable kind for 
study. Those of spermatocytes and odcytes are excluded from 
the survey, because in them the details are obscured by the 
complications due to the processes of conjugation. Most of 
the images described are from paraffin sections: surface 
preparations show nothing more than these. The most trust- 
worthy fixing agent has been found to be picro-formol (Bouin’s 
formula). Iron haematoxylin has been found to be comparably 
the best stain; but it should not be used quite as laid down 
in the books, which give excessive times and strengths. You 
should mordant (sections of 7-5 microns, or less) for not more 
than 2$ minutes in a solution of iron alum of 4 per cent. or 
