REPORT ON THE TUNICATA. 
151 
longer possible. Consequently, I feel compelled to arrange the large number of species 
composing the family in a series of closely allied groups which may be considered 
either as genera or as subgenera, and which are in some cases not very clearly separated 
from one another. The Polyclinidse exhibit peculiarly well the difficulties in classifying 
which are the natural result of the origin of species by evolution : most of the groups 
pass gradually into one another, and strict lines of demarcation are almost always absent. 
The subgenera proposed by Giard and by von Drasche have been retained, and along 
with the older genera of Savigny, Milne-Edwards, and others, and the new groups 
rendered necessary by some of the Challenger specimens, constitute the sets of species 
represented by the names in the Table below, and treated, for the sake of convenience, as 
genera in the following pages. In the Table (p. 152) the relative positions of the names 
do not always indicate the degree of affinity, and the distinguishing characters of the 
groups are stated in the briefest possible manner. The detailed characteristics and the 
genetic relationships are discussed under the head of each genus further on. I have 
avoided making use of the relative length of the post-abdomen as a distinguishing 
feature, as I am convinced that in some forms at least it is liable to very great variability, 
and even differs in size at different times in the life of the same Ascidiozooid. 
Three of the groups in the Table — Sigillina, Synoicum, and Sidnyum — are Savigny ’s 
genera unchanged. Two of them — Aj^lidium and Polyclinum — are genera founded by 
Savigny, but now used in a more restricted sense. 'EouT—Aurantium, Qircinalium, 
Frayarium, and Morchellium — ^were subgenera proposed by Giard. Amaroucium and 
Parascidia were established by Milne-Edwards, and Polyclinoides by von Drasche ; while 
the remaining five — Pharyngodictyon, Tylobrancliion, Atopogaster, Morchellioides, and 
Psammaplidium — are new groups founded for the reception of Challenger species. 
Pharyngodictyon is more clearly distinguished than any of the others, and is certainly 
worthy of distinct generic rank. In fact, it differs so markedly from all the other 
Polyclinidse in the structure of the branchial sac, that possibly it might be placed in a 
distinct family by itself. In regard to the others, it must be left an open question 
whether they are genera or subgenera. They are all tolerably closely allied, and some 
species present intermediate characters between two or more of them, thus rendering 
precise definition exceedingly difficult, if not imjDossible. On the other hand, it would 
be very inconvenient to regard the whole family, with the exception of Pharyngodictyon, 
as constituting a single enormous genus. I believe it will be found useful to retain the 
divisions as genera, at least until a much more complete knowledge of the species from 
all parts of the world of this variable and apparently very large family of the Compound 
Ascidians permits of a new and more natural classification. 
Circinalium, Giard, Fragarium, Giard, Parascidia, Milne-Edwards, Synoicurn, 
Phipps, Sigillina, Savigny, and Polyclinoides, von Drasche, do not occur in the 
Challenger collection, but the other eleven genera of the family are all represented. 
