VARIATION IN THE DESMIDIER. 395 
the forms he himself describes in the same paper as Staurastrum 
forficulatum var. longicorne (l. e. t. i. ff. 42-43); in fact, 
Schmidle’s S. aculeatum var. bifidum is, in my opinion, much nearer 
S. forficulatum, Lund., than is his S. forficulatum var. longicorne. 
The variations I have noticed in S. aculeatum are perhaps 
worth mentioning. The, most typical examples were from Capel 
Curig, N. Wales, and from the United States, the former speci- 
mens possessing the most strongly developed spines (PI. 11. 
fig. 31). From Orono, Maine, many examples were met with 
which had the spines considerably reduced in length, but were 
otherwise typical, and from Thursley Common, Surrey, a some- 
what smaller but fully developed form was abundant (Pl. 11. 
fig. 30). 
17. Sravrastrum vESTITUM, Ralfs, Brit. Desm. p. 143, t. xxiii. 
f. L—Few species exhibit so much variation as this, and at 
the same time retain their distinctive features. 1 need hardly 
mention that the main diagnostic character of 5. vestitum is the 
possession of a pair of furcate spines in the middle of the lateral 
margins of the vertical view. These spines are themselves 
subject to much variation, being sometimes simple aculei, at 
other times furcate to their base, and more rarely doubly furcate. 
The general plan of arrangement of the spines and emarginate 
warts on S. vestitum is precisely like that on 8. aculeatum. The 
angles of S. vestitum, which possess three well-marked divergent 
spines at their apices, are more produced than those of S. aculea- 
tum, and, as a rule, the two median spines of the dorsal 
series become converted into emarginate warts. Of the lateral 
series of spines, which are such a marked feature of S. aculea- 
tum, either the two median spines only remain in A. vestitum, 
or they are much more prominent than the rest. These are 
the characteristic furcate spines mentioned above. The front 
view of a typical form of 8. vestitum resembles very closely that 
of some forms of S. aculeatum, but in the majority of specimens 
of the former species the angles are produced into processes of 
various length. ln some these processes are very long (lat. c. 
proe. 90 uj), and in others they are very short (lat. c. proc. 46 po). 
In a gathering from Arderry Lough, W. Ireland, some forms 
were observed which in front view eould hardly be distinguished 
from S. anatinum, Cooke & Wills: long. s. proc. 35 a, c. proc. 
52 u; lat. c. proc. 90-98 pi; lat. isthm. 135 a. 
LINN. JOURN.— BOTANY, VOL. XXXIV. 2E 
