752 



SYSTEMATIC HISTORY OF THE rNTUSOEIA. 



2 * Endochrome i?i several spi?'al bands. 



S. ohscura (Ralfs). — ^Frond cylindrical 

 or fusifomi, five to eight times longer 

 than broad, extremities attenuated, ends 

 blunt ; endochrome in several slender spi- 



C. Cells 



Genus COSMOCLADIIJM (Breb.). 

 constricted at the middle, stipitate. 



CosMOCLADiuiM pulchellmn (Breb.). — 

 Stipes dendroid, dichotomously branched, 

 hyaline, with a slight intermediate thick- 

 ening between the cells ; cells terminal 

 and axillary, green, segments elliptico- 

 reniform, smooth (iii. 63). F. 



We here provisionally place this re- 



7'al hands, their revolutions two or three. 



! at each extremity, in which there is 

 sometimes a free granide. L. 1-247" to 

 1-226"; B. 1-1020" to 1-907". G.B., 

 I., F. 



sti/pitate. 



— Cells rounded, compressed, deeply 



markable plant, discovered by M. do 

 Brebisson, not knowing as yet anything 

 as to its mode of growth or development. 

 The ceUs, if detached from the stipes, 

 woidd scarcely be distinguishable from 

 those of Cosmarium bioculatum. 



D. Cells aggregated into families, forming fasciculi or faggot-like bundles. 



Genus AKKISTBODESMrS (Corda).— Cells minute, smooth, elongated, 

 attenuated, aggregated into families forming fasciculi or faggot -like bundles, 

 each family resulting from the self-division of a single cell, wMch commences 

 by the formation of a somewhat oblique septum at the middle, eventually 

 rendered more and more oblique from the young cells growing alongside one 

 another longitudinally until they each attain the length of the original parent- 

 ceU, the process being again and again repeated by each till the aggregated 

 family consists of at most thirty-two cells, the family finally again breaking 

 up into single cells. No other propagation known ; the position of the genus 

 is therefore doubtful. 



Ankisteodesmus falcatus (Ralfs). — 

 Cells very slender, arcuate (rarely straight 

 or sigmoid), graduallv attenuated, ends 

 acute. L. 1-550"; B. 1-7353". (i.35,36.) 

 = Rhaphidium fasciculatu?n (Kg., Nag.). 

 G.B., I., F., G., Italy. 



A. convoliitus (Corda). — Cells much 

 curved, crescent-shaped, somewhat ra- 

 pidly attenuated, ends subacute. = -R/i«- 

 phidium minutum (Nag.). I., F., G. 

 We have met with a plant (gathered 

 near Dublin) which we now (though 

 doubtfully) refer to this species, in which 

 we noticed self-division of the cells, in 

 an at first oblique, finally longitudinal 

 manner, very much the same as that 

 described by Nageli {Einzell. Aly.) for the 

 preceding species, and introduced into 

 the generic character. The cells in our 



plant are not quite so much curved as in 

 Nageli's drawing of this species, and are 

 rather more acute at the extremities : 

 we have not noticed the fascicidi to be 

 composed of more than 8 cells, frequently 

 of 2 or 4 ; and while so combined the cells 

 all look in the same direction, the con- 

 cave sm-face of the one being applied to 



I the convex sm'face of its neighbour. 



; A. contortus (Thuret). — Cells slender, 

 arcuate or sigmoid, somewhat gently in- 



! flated at the centre, ends drawTi out long 

 and very fine. F. 



[Scenodesmus duplex (Ralfs) is placed 

 in this genus by Kiitzing and Nageli 

 under the name of Rhaphidium ; that 

 plant may, however, be the cell of an 

 Ankistrociesmus undergoing division.] 



SubfamHy PEDIASTRE^ (page 24). 



We shall not attempt to give anything but a very provisional diagnosis of 

 the genera here included under the above title (which have long been asso- 

 ciated with the Desmidiacese, and chiefly for that reason finding a place in 

 the present work), as, so far as we can judge, it is not yet determined whether 

 they should remain united with the Palmellacece, to which they have been 



