25G 
THE VOYAGE OF H.M.S. OH ALLEN GEE. 
well in the Diclemniclse as in the allied family Diplosomidae, in which also pyloric 
gemmation is found in a very complete condition. 
The family Didemnidse contains only a small number of genera but has a consider- 
able number of species, which are in some cases very difficult to distinguish from one 
another. Savigny founded the two original genera, Didemnum and JEuccelium, in 1816.^ 
In his second memoir he characterises Didemnum as having both apertures of the 
Ascidiozooid superior (or anterior), the one with six regular lobes and the other irregular 
or simple ; while he places Euccelium in a distinct division, said to have both the aper- 
tures superior and simple. In his systematic arrangement, however (pp. 194, 195), he 
places the two genera close together, and defines them with more detail. The most 
important distinctions which he points out are that Euccelium forms a thinner and 
more delicate crust than that of Didemnum, and that the branchial aperture in 
Euccelium is circular and destitute of lobes, while that of Didemnum is surrounded 
by six equal lobes. He describes two species of Didemnum, viz., Didemnum 
candidum, and Didemnum viscosum, both from the Gulf of Suez ; and one species of 
Euccelium, Euccelium liospitiolum, from the same locality. Didemnum candidum and 
Euccelium liospitiolum are figured {loc. cit., pi. xx.), and from these figures some 
further distinctive features of Savigny ’s two genera can be made out. In the external 
appearance the only noticeable distinction appears to be that the cloacal apertures 
(which Savigny regarded as being absent, but which are distinctly visible in his 
figure^) are much more obvious in Euccelium than in Didemnum. In the minute 
structure there are several points of distinction. 1. The spicules in the test of 
Didemnum are almost spherical, having merely very slight projections, while those 
of Euccelium have much larger projections so as to present more of a stellate 
appearance. 2. The pedicle connecting the branchial and visceral regions of the body 
is longer in Didemnum than in Euccelium. 3. The stomach is on the dorsal side of 
the intestine in Euccelium, while it is on the ventral side in Didemnum. 4. The 
rectum reaches almost to the anterior end of the branchial sac in Euccelium, while in 
Didemnum it terminates about half-way forwards. 5. As Giard (see below) has pointed 
out, the mantle in Euccelium is thin and transparent, allowing the branchial sac to 
show through, while in Didemnum it is thick and opaque. 6. Euccelium has a peculiar 
little swelling upon the posterior part of the intestinal loop behind the stomach which 
is not found in Didemnum. These characters, however, do not apply to all the species 
of Didemnum. 
Savigny’s two genera, as represented by the type species which he figures, Didemnum 
candidum and Euccelium liospitiolum, seem to be well characterised and sufficiently 
distinct from one another. Lamarck,^ however, writing immediately after Savigny, seems 
to have tliouglit otherwise, as he united the two genera under the one term Euccelium. 
^ Memoires, p. 64, and Systeme, pp. 194, 195. ^ ^.oc. cit., pi. iv. fig. 4. ^ Hist. Nat, &c. 
