GROWTH. ETC.. I\ THE DESMIDI.A.CE.E. 289 



as if they had been formed from the old nucleus by division, but 

 as if they were of separate growth. I see no reason why secondary 

 pyrenoids should not form de novo as well as the original ones. 



In the long Closteria we may often see one series of very 

 distinct pyrenoids which alternate with fainter or smaller ones 

 in a second series. Sometimes the pyrenoids are so close together 

 as to be almost touching. The second series, I have noticed, is 

 usually on a slightly different plane in the cell to the other, and 

 it may well be therefore that Kleb's figure of CI. lunula (3 coloratiim, 

 T. i, f. ib, is CI. lanceolatmn forma with two central rows of pyre- 

 noids. Such I have considered it in " Some Sydney Desmids." 

 As the curved Closteria grow in length an extra pyrenoid, or more 

 than one, often forms towards the tip of the semi-cell. 



Since wTiting the above a fine gathering has been obtained from 

 a small pool at Canley Vale (110 N.H.S.)^ in which Cos. psendopro- 

 tuberans was in profusion. The endochrome was in excellent state 

 for observation and presented a variety of conditions. Some cells 

 had the chloroplasts plainly divided vertically, but the pyrenoid 

 was single and central. In others again the pyrenoid, while single, 

 was evidently dividing vertically, the centre from top to bottom 

 being much more faint than the sides, and the shape somewhat 

 produced laterally. Other cells of this variety had indifferently 

 one or two pyrenoids to the semi-cell, the number of cells with each 

 being about equal. Here and there also one might be seen with a 

 single pyrenoid in one semi-cell and two in the other. It seems 

 certain, then, that my suspicion was well founded, for I am certain 

 that Cos. pseiidoprotuberans, Kirchn., is the penultimate subspecies 

 (or variation) of Cos. rectangiilare, Grun., which always has binate 

 pyTcnoids. It is evident also that the number of pyrenoids has no 

 value at all as an indication of specific identity or for 

 purposes of classification. The pyi'enoids double in every species 

 at latest in the penultimate form, and no Cosmarium can therefore 

 be called mature which has not geminate pyrenoids to the semi-cell. 



Endochrome. — As the cell develops, so do the chloroplasts. 

 The single flat axillary' chloroplast found in Gonatozygon and 

 certain Closteria does not always remain flat. A longitudinal ruck 

 forms down the centre on each side and it becomes an axillary four- 

 rayed chloroplast. I have observed this arrangement in Gon. 

 monotcenium var. Brebissonii, rolling the cell over several times to 

 make sure. Much the same development is found in CI. Leihleinii. 

 In the same gathering I have noted specimens with a simple axile 

 lamina, also with four and with six longitudinal plications. The 

 third and fourth are formed as longitudinal plications down the 

 centre of the original lamina, and later they seem to give rise to 

 the next pair. As the cell develops the number of ridges increases. 

 The same is true of the parietal chloroplasts of Docidium (including 

 PleurotcBnitim) . In the smaller forms of Doc. trabecula, for ex- 

 ample, such as var. Ehrenbergii, there are only four chloroplasts ; 



1 Samoles deposited in the National Herbarium, Sydney. 



