252 SUPPLEMENT 
The general study of this form then gives a division into 
those which are tubular and those which are not. 
We call tubular shells those whose transverse diameter is 
considerably smaller than the longitudinal, and which are 
not enrolled or turbinated, or if so, only in a very irregular 
manner, and never in a spiral. ‘These are the tubes of certain 
genera of setépodes, which have another distinctive character, 
namely, that the summit. (head) is always open, which is 
never the case with the shells of the mollusca proper. 
The non-tubular shells are then divided into shells of a 
single piece, these are the wntvalves; and into shells of many 
pieces, or multivalves ; and these last into bivalves and multi- 
valves, or dissivalves. 
According to this, we must understand by valve (valvula) 
a calcareous piece of a very variable form, applied on or in 
the skin of a molluscous animal, and covering a greater or 
less part of it; but then we must often have recourse to the 
skin of the animal to judge whether a certain number of these 
valves belonged to one individual; as, for example, when 
they have no direct relations between themselves, but only 
indirect by means of the skin. ‘This is the reason why for a 
long time one valve of the testa of the lingula has been re- 
garded as an univalve shell. 
The multivalve shells are of three sorts; those which are 
composed of many transverse imbricated pieces, as in osca- 
brio, those which are formed of five or many valves, sym- 
metrically ranged on the right and left, and sometimes even 
placed in scales, and united together by means of the skin, as 
in anatifa, (these are the dissivalves of M. Denys de Mont- 
fort), and finally, those which are disposed in a manner almost 
circular, as in the balani and neighbouring genera. 
The bivalve shells are those which, as their name indicates, 
are formed only of two pieces; sometimes, it is true, enclosed 
‘ 
ee 
