432 Dublin Natural History Society. 



have referred in the section of the genus alluded to. The number of 

 spiral tubes varies in this and other species in different filaments, but 

 not in the same, and this makes a corresponding difference in the 

 length of the joints or cells, which are longer if there be four or five 

 tubes instead of three. The proofs now to be adduced, that this mode 

 of growth likewise takes place in ail Confervse which are composed 

 of simple unbranched filaments, a large class, are little less conclu- 

 sive than those just enumerated. In most of the filaments of these 

 the cells will be observed to be of various lengths, some twice as long 

 as others, and these again of every intermediate length. Now, by 

 means of this law of growth, this variation in the length of the cells 

 is at once and satisfactorily accounted for, which is not to be done 

 in any other way. But this is not all^the progress of the formation 

 of the septa which divide the cells may be frequently traced either in 

 the same or different filaments, Avhich is alone sufficient to establish 

 the reality of the existence of this law of increase in this numerous 

 section of the class Conferva. The only Confervas to which I should 

 for a moment hesitate to apply this method of development, and I be- 

 lieve that it is apjslicable to them likewise, are the branched species, 

 to wdiich such a means of increase is less necessary, seeing that, un- 

 like those with simple unbranched filaments, they have innumerable 

 terminal points of growth. Now I beg to lay particular stress on 

 this law of development, which is evidently very important, inasmuch 

 as it not merely goes to account for the rapid growth of many spe- 

 cies of Confervse — for it is simultaneously in operation in each of the 

 many hundred cells of which each filament of most Confervse is com- 

 posed — but it likewise teaches us that much caution is requisite in 

 determining species, as it proves that the character most relied on 

 for this purpose is one subject to verj'- great variation — that is, the 

 length of the joints. There is a limit, however, to this law of deve- 

 lopment, which does not, in the section of the genus Conjngata to 

 which reference has been made, allow of more than one or two divi- 

 sions of each cell, unless, indeed, the spiral tubes grow likewise in 

 an equal ratio, which may be the case, and then the division of the 

 cells may be frequently repeated. In those Confervse which do not 

 contain spiral tubes, the multiplication of the cells may go on to an 

 almost endless extent. To illustrate the importance of attention to 

 this law of development in determining species, I may observe, that 

 but for its timely discovery I should have described several species 

 of Conjiiyata as distinct, which are really not so, considering the 

 length of the cells and number of spiral tubes in the interior of each 

 cell to be the most decided characters whereon to found specific dif- 

 ferences. They are not so, however, one of the most certain being 

 the diameter of the filaments. But carrjdng this law in view, it is 

 not difficult to estimate the extent of the variations in length to which 

 the cells are subject, first ascertaining what the primary length of the 

 ceU is. In the branched Confervae there are laws of development, 

 some of them peculiar to each species, presiding over the arrangement 

 of the branches and cells, which have hitherto escaped the scrutiny 

 of man." 



