On the Rhizocephalan Genus Thompsonia, etc. 9 
Decapod Crustacea respectively, although, owing to the scantiness 
of my material, I was unable to investigate thoroughly those of the 
latter class. The account which follows is based almost entirely on 
the parasite of Synalpheus. 
THE ROOT SYSTEM. 
The root system is the characteristic organ of the Rhizocephala. 
Delage and Geoffrey Smith have shown that it appears first in the 
course of endoparasitic development and is well established before the 
visceral mass begins to form. There is no doubt as to its existence 
and importance in all genera, with the exception of the doubtful form 
Duplorbis, but it has been exactly described only in Sacculina. The 
following brief description is condensed from the account of Delage: 
The tissue of the peduncle is bounded by the basilar membrane and from 
this the roots originate. For the most part they originate in a small number of 
large trunks which immediately ramify and spread. At first the roots follow 
the intestine, which they envelop with a close network. Some then pass into 
the distal part of the abdomen of the crab, penetrate the muscular layers of the 
body wall, and enter the abdominal appendages. Others pass forward and 
divide into two groups: the first follows the intestine and passes on to the 
stomach and into the interstices of the liver lobes and the genital glands; 
the second, not quite so important, accompanies the nerve chain, its branches 
following the lateral nerves into the limb muscles and penetrating into all 
the appendages, limbs, antenne, and even the peduncles of the eyes. All the 
organs are invaded except the gills and heart. 
The roots branch with an irregular dichotomy. The separate branches 
never anastomose. In diameter they usually vary from 30 to 40u, but the 
largest may measure 200u. Histologically they consist of the following ele- 
ments: (1) an exceedingly thin external layer of chitin; (2) a layer of cells 
forming a complete internal investment; (3) the interior of the root is occupied 
by star-shaped cells, the processes of which anastomose with each other and 
with the marginal eells. In the largest roots the centre isempty. Refringent 
drops or granules of yolk occupy the meshes. The root system of Peltogaster, 
though similar to that of Sacculina in details of histology and method of 
branching, has a very different distribution. It constitutes a compact mass 
which is concentrated in the neighbourhood of the peduncle, not invading the 
spaces between the tissues of the host in the manner characteristic of Sacculina. 
Such accounts of the root system of Thompsonia as have been given 
by the earlier observers are conflicting and obviously influenced by a 
predisposition to accept the theory that each external sac is a meta- 
morphosed larva. It is, moreover, exceedingly difficult to demonstrate 
the root system of Rhizocephala when preserved by usual methods, for 
the roots, owing to the impenetrability of their cuticular envelope, take 
up stains much less readily than the surrounding tissues. In addition, 
the highly refringent yolk globules, which make the roots easily dis- 
tinguished in life, disappear under the action of alcohol or even formalin. 
And lastly, the roots are even more delicate than in Sacculina, so that we 
