363 WILLIAM ARCHER. 



tuft of hair-like filaments surrounding the terminal '^ hetero- 

 cyst" of a Cylindrosperum may be something analogous 

 But there they are irregular and more hair-like ; here solitary, 

 definite in position^ longer, and more like a very delicate 

 linear spine ; that is, not thicker at the base, or at any portion 

 of ite lenarth. I have met with this oro-anism on several 

 occasions in moor-pools, chiefly in the counties of Dublin 

 and Wicklow ; it is possible it may be pretty generally 

 diffused, though not often seen, which may, indeed, in great 

 part be due to its great minuteness. Triradidte and setigerous 

 are thus the special features of this form. 



Having thus tried to convey a conception of these four 

 unicellular figured chroococcoids, the two first mentioned of 

 which form Reinsch's genus Tetrapedia, as established by 

 that author (op. cit.), it remains to consider in what manner 

 the two latter, hitherto unrecorded, could be associated there- 

 with. Taking the terms of the genus Tetrapedia, as founded, 

 the two new forms would not correctly fall thereunder, the 

 characters being based necessarily on the outward figure. 

 If, indeed, it should be rigidly held that my two forms 

 should be decidedly excluded from Reinsch's genus, in such 

 case, fixing the generic limits on the same principle, each of 

 these would seem to demand to be regarded as the types of 

 two separate new genera ; for to myself each would appear to 

 be as distinct from each other as either from the two 

 Reinschian forms. But, nevertheless, all the four forms are, 

 I venture to suppose, sufficiently clear to be regarded as 

 closely kindred ; they are all compressed, angular, deeply 

 subdivided either by narro^v, even linear incisions, or by broad, 

 angular, or rounded sinuses. 



The third form referred to (figs. 11, 12), with its two 

 broadly cuneate segments, is com^jaraS/e to T. Crux-Michaeli, 

 as it Avere, deprived of two opposite segments; no doubt, 

 even for a moment supposing it possible to operate on an 

 example so as actually to remove two of its segments, the 

 result would still not be at all identical with my form in the 

 contour of the remaining segments ; therefore, so far as one 

 can fairly judge, it would not be a reasonable assumption to 

 regard it as a binate or, so to say, depauperated variety of T. 

 Crux-Michaeli (Reinsch). It does seem the new form alluded 

 to is not apparently at any time, so far as it has presented 

 itself, divided " incisuris quaternis," yet still the approxima- 

 tion or " affinity" cannot but be regarded as sufficiently 

 striking. The fourth form (figs. 14 — 16), with its three 

 angles only, and its slender filiform or bristle-like processes, 

 appears, indeed, more distinguished from the rest; but apart 



