KONGL. SV. VET. AKADEMIRNS IIANDLINOAR. HAND 26. N:0 2. 47 



first, the presence of longitudinal lines and secondly, the striae, which are not distinctly punctate, 

 the stria- of all tlie species of Pleurosigma being punctate. 



An asymmetrical tendency is slightly apparent in several forms of C. Liher, and is very decided 

 in (J. (AUnioyieis) curvinervis. Among the Amphora' are many forms with longitudinal lines, but 

 no known species of that genus is closely connected with Colonels. — There are vai'ious transitions 

 among the forms of Caloneis. Some species are slightly constricted in the middle or biconstricted, 

 and thus connect the non-constricted forms with the panduriform or strongly biconstricted. 



('lo.sely connected with Caloneis is the genus F.-^cudodiiipliiprora, among the species of which 

 F. itnplcta comes nearest. The principal distinction between these two genera consists in the na- 

 ture of the striii", which are punctate in Fscndoa ntpliiprora . The transverse expansion of the cen- 

 tral nodule of Pseudo-amphiprora is met with in some species of Caloneis, as for instance ('. Wardii. 

 — The terminal fissures in most forms of tlie type C. Liher are not very divergent from the 

 direction of the median line, and being situated on the concave end of the valve, thpy have the 

 appearance of short flexuose lines (see A. Schmidt's Atl. PI. L.). In C (ihuormis they are turned 

 in contrary directions. In the forms of the type C Poivellii and in tlie panduriform species they 

 are small and indistinct. 



The striie are as a rule transverse, in some forms slightly radiate, and are usually di^■ergent 

 at the ends, not convergent as in Pinniilaria". They are generally smooth or not distinctly punctate, 

 excepting in the doubtful C. (lisj)ersa, whicli has distinctly punctate strite. — In addition to the 

 nature of the striae, the longitudinal lines are a very characteristic feature of f-aloneis. In many 

 species the valves are divided by these lines into an inner plane, or even concave, portion; and an 

 outer portion which is convex and sloping. In other species the inner poi'tion ajjpears to be more 

 elevated. W'itliout transverse sections of the valves it is extremely difficult to ascertain which 

 is the case; but as a rule the plane, or inclination, of the part of the valve which is within the 

 longitudinal lines is different from that of the outside jmrts. Transitions of all kind soccur, from the 

 narrow true lines of C Liber and others to the broad bands of C Castraceniei, C. mirahilis etc., 

 shewing that, whether narrow or broad, these lines are of the same nature. — In C. Madagascarensis 

 there are, in addition, otlier fainter, longitudinal lines, formed by knots on the stria^; and this 

 fact seems to point to an affinity between Caloneis and Tropidnneis (T. lepidoptera, T. longa). 



I" 



Artificial hey. 

 f'entral nodule stauroid C. Wnrclii C'l,. 



not — • 2 



q I Central area a transverse fasria ;3. ') 



"■ I — — not — — 19. 



„ ^ Valve sigmoid (\ sfnurophora Grtjn. 



( Va: 



straight 4. 



. j Linear or almost linear 5. 



\ Lanceolate, constricted or biconstricted 11. 



_ J Striae radiate ('. Clevei Laost. 



' \ — parallel (j. 



,. I Axial area rather broad 7. 



X). < 



1^ — — narrow or indistinct 8. 



_ j Marine species C. lafrfasnafa Urun. (('. formosa var. interriipta). 



\ Freshwater species C. patngonica Cl. 



J Margins parallel 9. 



I — slightly undulate C. Siliciiln var. 



, J Stria; 1:') in O.ni mm C. consimilis A. S. 



' ■ 1 — finer 10. 



' To this division belongs f. nnhirnJn ("tRun., which for vant of figure and complete description cannot be enrolled 

 in the kev. 



