432 Mr. C. Mereschkowsky on Stauronella. 
Breadth of At the 
Length. the frustule. constriction. 
mm. mm. mm. 
0:0295 Bie Bs 
0:0295 0:0076 0:0057 
0-042 Hees ite 
0:0437 0:0095 0-0067 
00457 “sons agit 
0:0445 0:0086 Lae 
00466 0:0095 0:0076 
0-0466 ee ein 
0:0475 
0:0475 Jawa atone 
0:048 0:0086 0-0067 
0-049 ie Baca 
0-050 0-0105 0-0086 
0-050 See: Shyee 
0:052 0:0105 - 0:0086 
0:055 Pore oe yi 
0-056 0:0105 0:0086 
Average .. 0:0413 0:0092 0:0073 
Var. linearis, Mer., nov. var. (Pl. VIII. figs. 7-19.) 
Diagnosis.—Valve narrow, linear or slightly attenuated at 
the ends, which are broadly rounded, never constricted in the 
middle. Girdle-face rather broad, quadrangular, slightly 
constricted, with round angles; extremities broadly truncate. 
Divisions of the zone terminating in marginal puncta, the 
four outer ones being very conspicuous. Length 0:015-0-053 
mm. (average 0°037 mm.) ; breadth of valve 0:0048-0:0067 
mm.; breadth of girdle-face 0°0095-0°012 mm., at the con- 
striction 0:009-0:0095 mm. 
Locality. Le Havre, North France (mar.); Black Sea, 
Crimea (lalta, Sebastopol, somewhat brack. and mar.). 
Of this variety I have seen a very great number of indi- 
viduals in the Black Sea. In one gathering from [alta 
(Crimea) I found it in an almost pure condition, with only 
occasionally a few specimens of the type species ; in another 
gathering from Sebastopol it was the type species which 
predominated, only very few individuals of var. Anearts being 
mixed with it. This shows that the var. linearis is really 
distinct, while intermediate forms prove it to be only a 
variety. 
The valves are sometimes perfectly linear (figs. 10, 17), 
but as frequently, if not more, they are somewhat attenuated 
at the ends (figs. 7, 11, 15), the extremities being sometimes 
a little produced (figs. 13, 16) ; the ends are not truncate, as 
