160 FRANK W. BANCROFT 
pure F. heteroclitus this chemotorpism of the chromatophores for 
the blood vessels is destroyed by the stoppage of the circulation 
the black chromatophores of this form also develop more branches 
than usual and may even in extreme cases (such as fig. 24) simu- 
late closely the chromatophores of the pure F. majalis. This 
instance shows very clearly the importance of an analysis of the 
mechanism of heredity insisted upon by Loeb in 1898* and more 
recently by Newman. 
3. Head chromatophores 
One of the factors which influenced Newman in arriving at 
the conclusion that in the hybrids between majalis and hetero- 
clitus, blended inheritance predominates is the manner of the 
appearance of the head pigmentation. Newman found, and I 
have confirmed his results, that the head pigment appears first 
on the pure F’. heteroclitus; next on the F. heteroclitus egg hybrid; 
next on the F’. majalis egg hybrid; and last on the pure F. majalis. 
Now, while this is true, still a closer analysis of this process has 
furnished what is probably the clearest case of Mendelian domi- 
nance encountered in this whole study. 
In the pure F. heteroclitus, and in both hybrids the black head 
chromatophores appeared in two crops separated by an interval 
of about two days. In the pure F. majalis, on the other hand, 
this first crop was wanting and the second crop appeared at about 
the same time that it did in the pure F. heteroclitus. 
The first head chromatophores appeared on the first or second 
day after the first appearance of the yolk chromatophores. They 
were first seen on the sides of the brain, having probably migrated 
in from the yolk, and wandered onto the dorsum of hind and mid- 
brain, only in rare cases reaching the fore-brain. Various stages 
in the development of these cells are shown in figs. 1 to 6. As 
soon as they reached the dorsum of the brain they began to expand 
and finally became very conspicuous objects (fig. 5). Figures 
2 and 3 which show these chromatophores before and after an 
interval of three and a half hours give some idea of the rapidity 
with which this migration and expansion takes place. 
6 Marine Biol. Lab. Lectures for 1897 and 1898, pp. 227-229. 
