228 ARCHER, ON DESMIDIACEJB. 



the new addition is an extension to the base of one only of the 

 original segments. The growth of the additional lateral tube 

 in the present instance is comparable rather to the somewhat 

 similar extension from the shorter or younger cone, prepara- 

 tory to conjugation, in Closterium EArenbergii, described a 

 little back, except that here it is usually longer (or more than 

 one), and gives egress to separated portions of the endochorme, 

 individualized as zoospores, and not permitting it to extrude 

 en masse for the purpose of conjugating with the contents of 

 a neighbouring frond ; thus Ave not unfrequently observe in 

 nature modifications of similar means conducive to different 

 ends. When the segment gives rise to one tubercle only, this 

 additional growth is gradually developed more and more 

 narrowly, diminishing to nothing at the opposite point of the 

 cylindrical segment, so that the frond is thus thrown out of its 

 straight or nearly straight direction, and becomes bent into 

 a knee-shape (Figs. 2 and 3). Such is also the case when 

 two projections arise side by side. But when two originate 

 opposite to each other, or when there are three, the frond is 

 not thrown out of its straight form, because the new addition 

 to the segment, from which these lateral growths take their 

 origin, now forms an annular extension equal all round, and 

 the segment therefore becomes added to in length by just 

 so much as the annular addition is broad — and this is less 

 than the 1 -3000th part of an inch (Fig. 4). As the case is 

 pretty similar whether there be one, two, or three of these 

 lateral growths, I shall continue my remarks upon those cases 

 where one only is formed. The basal tubercle now gradually 

 elongates, and becomes a tube in direct connection and con- 

 tinuation with the frond (like the finger to a glove), and is 

 about l-3G00th of an inch in diameter, but of very varied 

 degrees of length (Fig. 2). I have noticed some to cease to 

 grow after having barely attained about l-10th or l-8th part 

 of the length of the frond, and I have seen a few very Long, 

 almost, if not quite as long as the frond itself. The endo- 

 chromc near the base of each segment, and in the neighbourhood 

 of the lateral tube, next becomes very finely granular, of an 

 almost homogeneous appearance, and the lateral tube is filled 

 by it. The remainder of the endochrome (even in the state 

 indicated by Fig. 2) is still but little altered from the ordinary 

 condition, and the terminal cavities with the active granules, 

 characteristic of this genus, as well a of Closterium, remain 

 unchanged. The annular addition and the lateral tubes 

 are quite smooth, and destitute of the scattered puncta 

 which characterise the emptv frond in this species (Fig;;. 

 3, 4). 



