64 Proceedings of tlie Royal Society of Edinburgh. [Sess. 
genus Plakarthrium (Sphaeromidae), the coxopodites of the second thoracic 
somite (the first free somite) are completely coalesced with the body.” 
The trend of the present discussion will now be obvious. It is that the 
moult unmasks the line of junction between fused coxopodite and body- 
ring in the case of the second to the seventh (free) somites ; in other words, 
that the longitudinal splitting is after all inter-segmental. 
As proof I adduce the following facts. In Idotea baltica and in 
Idotea emarginata the line of junction between coxopodite and tergite (in 
the case of segments 2 to 7) is visible dorsally. During moult a longitudinal 
split occurs at these junctions, no corresponding split occurring in the first 
(free) segment. In Asellus aquations , in which the coxopodites do not form 
coxal plates, the body-rings are detached complete. Finally, in Amphipoda 
( Gammarus , Orchestia, Talitrus), in which the moult is not a double but a 
single event, splitting occurs along the line of the coxotergal junctions of 
all ambulatory limbs. 
Examining the dorsal aspect of Ligia, one notes a faint line on each side 
running backwards from the animal’s eye parallel to the lateral border. 
On the anterior four (free) thoracic somites this line is composed of two 
parallel furrows separated by a very narrow ridge ; on the remaining three 
segments it is represented by a single furrow. When Ligia moults, the 
three posterior pairs of coxal plates separate from the tergites at the single 
furrow ; the three preceding pairs separate at the more lateral of the 
two parallel furrows. The coxopodites are then found to have a convex 
upper border, while the tergites are hollowed out (as in some Idoteae) to 
correspond. In those Oniscoidea in which the line of junction between 
coxopodite and tergite is not externally marked, e.g. Oniscus , Porcellio, 
Philoscia, a similar separation, disclosing similar moulding, occurs on 
moulting. 
One or two points of general interest may be mentioned. In retaining 
traces of the original line of junction between coxopodite and tergite Ligia 
is more primitive than many Oniscoidea. The absence of splitting between 
the coxopodite of the first walking limb and its corresponding tergite gives 
interesting support to the conclusion (otherwise derived) that union of 
first coxopodite with body occurred early in the ancestors of Isopoda. The 
occurrence of coxotergal splitting in amphipods as well as isopods, and its 
absence in decapods, shows that this mode of moulting is an important 
arthrostracan, if not peracaridan, feature. 
Transverse Splitting. — We should now be more inclined perhaps to 
attribute systematic importance to the site of the transverse split. Accord- 
ing to Lockwood (1870) the cuticle of Limulus splits along a “thin narrow 
