291 
distinct surfaces externally, as shown by the accompanying 
figure ; a, the dorsum, gene- 
rally more or less concave, 
and b, b, the two lateral sur- 
faces. It will be seen, there- 
fore, that the margins (c, c) 
are contracted within the area 
of the expanded alse (rf, d) 
which are composed of two 
folds of the valve, and that the 
cariruE ( e , e) are the points of 
flexure. 
The transverse or oblique “ canaliculi ” of Smith, or 
“ costce ” of Rabenhorst, and others, are the alternate 
plicce produced by the corrugation laterally of the valve. 
The plicae are alternately concave and convex on both 
surfaces. In the typical species in which the plicae are semi- 
cylindrical — as in S. biseriata and S. splendida — that which 
is a costa on one surface of the valve is a canaliculus on the 
Ideal transverse section of a 
valve of Surirella. 
opposite. The object of this structure appears to be to secure 
sufficient strength and a certain degree of elasticity, with a 
small amount of siliceous matter, so as to enable the frustule 
to float in water. 
When a valve is placed flat under the field of the micro- 
scope the margins are overlapped by the carinse, and are thus 
concealed from view. 
1. Surirella Barrowcliffia, Donkin. 
Frustule panduriform on M. V. 1 and S. Y. (and when 
placed obliquely), with four equidistant, expanded, carinated 
alse ; carinse sinuously constricted in the middle and rounded 
towards the truncated extremities. Valve on S. V. elongated, 
narrow, slightly panduriform ; extremities rounded with a 
central notch ; alse expanded backwards, panduriform, 
strongly carinated ; plicse (canaliculi) transverse, wide, semi- 
cylindrical; widening towards the carinse and reaching a 
short distance within them. 
Hub. Fresh water. River Coquet, Elyhaugh ; plentiful 
April and Wav, 1869. 
The panduriform outline of the species on every aspect 
1 F. V., Smith. I prefer M. V., to denote marginal view, or that view of 
the frustule in which the margins of the valves with the intervening connect- 
ing zone are exposed to view. Great confusion in this respect prevails. 
What is a primary surface with Kutzing, is a secondary with Rabenhorst, 
and vice versa. 
