28 University of Michigan 
laterals, the ventral being especially high and lamellate; right 
valve with two, parallel, oblique, usually lamellate pseudo- 
cardinals, although the upper is vestigial or sometimes almost 
completely lacking, while the lower in old shells tends to become 
almost trigonal; and with one thin lateral; beak cavities mod- 
erately deep, capacious, obscuring the dorsal pits anteriad; 
anterior muscle scars well marked, smoothish, separated ; pos- 
terior scars larger, confluent, not impressed, concentrically 
striate and iridescent; pallial line well marked throughout its 
length; nacre white, or tinted with lavender or salmon, some- 
what thickened anteriad and delicately iridescent throughout. 
The male(?) shell (plate VIII, figs. 39 and 40, type) is sub- 
rhomboid, with the posterior ridge better marked and ending 
in a rounded point about one-half way up on the posterior 
margin. The female(?) shell is more elongate, subovate, and 
strongly inflated in the posterior half of the shell, with the 
very much rounded posterior point one-half or more of the 
height above the ventral margin (plate VIII, figs. 41 and 42). 
This subspecies is apparently a smaller form of the more 
southern typical rovirosa. The adult female(?) shell also 
differs from the type specimen, by a tendency to be somewhat 
more strongly inflated and elongate. 
Simpson (1900) first pointed out that U. testndineus Reeve 
differed from true explicatus Morelet, and named the form 
L. lividus. Although I must confess that I am unable to place 
some individuals to my complete satisfaction, I think that the 
two shells, as Simpson intimated, are not even very closely 
related. But, from the material on which the subspecies 1s 
based, U. testudineus or L. lividus appears to represent a rather 
unpronounced female type, while rovirosai (type in A. N. 8. 
P.) is what I believe to be the completely developed, old female, 
which has a type of marsupial(?) swelling quite distinct from 
