252 Triassic Echinoderms of Bakony. 
need for fulfilling this function. Dr. J. von Urxkiit (1896) has shown that their 
extrusion and retraction are effected by changes of pressure on the fluid in the 
peripharyngeal sinus, and that this pressure is normally produced in life by the 
contraction of the «ligamenta obliqua externa» VaLentTin, which pass from the forked 
ends of the compasses to the interradial regions of the perignathic girdle. Thus 
we need no longer say of these muscles with Loven (1892, p. 43), «their function 
is not yet properly understood». On the other hand von UrxxktLt shows that the 
pressure is relieved by the contraction of the compass-muscles, that is to say the 
five broad muscles that join the sides of the compasses and form a pentagonal ring 
at the adapical end of the lantern; when this ring is raised up in a funnel-shape, 
the space of the peripharyngeal sinus is enlarged, and the fluid withdrawn from the 
external branchiae. The Cidaridae have compasses with their associated muscles, 
and are similarly subject to.a change of volume in the peripharyngeal sinus. This 
is perhaps not so marked, since the limiting membrane appears to be less flexible; 
still some provision has to be made for the reception of superfluous fluid, and 
reservoirs are duly provided by the organs of Stewart, sometimes called «internal 
branchiae». It is, however, also the case that the internal pressure is partly taken 
up by the peristomial membrane, as may easily be proved by direct experiment. 
The greater the extent and flexibility of this membrane, the less would be the 
mechanical necessity for either external or internal reservoir-sacs. In the older 
Echinoids with a flexible test and less rigid perignathic girdle, the peristomial mem- 
brane and its surroundings must have yielded to the internal pressure far more 
than in the rigid forms of later origin. In the Silurian Palaeodiscus, where no 
trace of a perignathic girdle has yet been described, and where the differentiation 
of plates in the peristomial region is very slight, the peristomial limits of the peri- 
pharyngeal sinus are readily distinguished in the fossils, owing to the protrusion 
of the plates on the death of the animal in consequence of the greater pressure of 
the contained fluid. In this genus, and probably in all primitive Echinoids, the 
lantern was not so high as in later forms, and the cavity of the sinus must have 
been relatively smaller. Consequently changes of pressure would not have necessi- 
tated large reservoir-sacs; and it may be that none existed at all. 
The facts that have been very briefly summarised in the preceding paragraph 
may perhaps permit some speculation as to the course of evolution. They suggest 
that, as the test became more rigid, especially in the circumoral region, as the 
lantern and its containing sinus increased in relative size, and as auricles and ridges 
were formed on the interambulacra for the attachment of the lantern-muscles, so 
there arose the need for reservoirs. These at first need not have been due to 
respiratory requirements, and in this regard it is suggestive to note von UExkuLL’s 
experimental proof that Cidaris papillata is far less subject to asphyxiation than 
the ectobranchiate genera which he investigated. These reservoir-sacs were, we 
know, developed in two different places: first, the organs of Stewart, passing up 
from the compass-sacs into the general body cavity; second, the external branchiae. 
The former, though called branchiae, cannot, one would think, subserve respiration 
to any great extent; but here further experiment is wanted. The latter, as soon 
as they reached any size, must have been far more effective in respiration than the 
organs of Stewart, owing to their contact with the external medium. It may be 
that the Permian and early Triassic Cidarids were still halting between these divergent 
