86 
Proceedings of the Royal Society of Edinburgh. [Sess. 
mitotic figures are present. The thickening is most marked in the centre, 
and it presents the appearance of a raised disc. 
By the twelfth day the cavity between the new diaphragm and the 
layer proliferated across the stump has changed its character. It is now 
dilated and the outer layer is bulging (fig. 8). The inside of the hemisphere 
so formed is invariably found after fixation filled with clear coagulated 
plasma. At the upper free border of the ventral flap of the diaphragm the 
nerve is found to be growing into the cavity, as is also the artery, though 
it may be here mentioned that the latter passes in at a later date than the 
former. The nerve, which from the beginning had its end in contact with 
the central part of the new stump-membrane, retains itself in this condition, 
and as the cavity increases in length it proliferates, sending out cell-chains 
which run in various directions. Some form a lining on the inside of the 
new epidermis, and others pass directly into the extreme tip of the new 
papilla, which is now the main centre of growth. Reed (6) holds that cells 
bud off from the main ectoderm and, passing to the free nerve-end, are 
added on to it, and in this way the nerve increases in length and comes 
to lie in the cavity. I am unable to agree with this, however, and from 
my observations am convinced that the nerve follows by its cell-chains 
the centre of proliferation with which it is always in contact. Further 
evidence is that any injury to the papilla, either by cutting or puncturing, 
is always responded to by violent movement on the part of the crab. This 
would indicate that in its new member sensory innervation seems to be as 
complete as in the other legs. Such a condition exists from the beginning 
of growth, and cannot therefore be explained by a secondary junction, as 
Reed seeks to prove. 
At the centre of proliferation at the tip of the papilla, mitotic figures can 
be seen scattered along the outer parts of the cavity wall, and they are most 
abundant at the tip. Here proliferation of two kinds occurs. Many cells 
remain in the wall after division, and take up their places as part of the 
epidermis. Others are seen in little groups or nests of five or six. These 
groups pass into the cavity free from the walls and come to lie in masses 
(fig. 9). The pink-staining cytoplasm of each group becomes collected in 
the centre, and fusion of the cells seems to take place, the nuclei lying on 
the outside edge of the mass. As the growing point recedes from the base, 
these cell-masses are deposited across the cavity, and later become differ- 
entiated into muscular tissue. 
By the twentieth day (in crabs inhabiting ordinary periwinkle shells) 
constrictions can be seen on the outer side of the papilla, and the segmented 
form thus presented imitates, in a way, the normal limb. Microscopic 
