IGO HICKS^ ON THE DIAMORPHOSIS OF LYNGBYA, ETC. 



for the cell- division to extend laterally instead of linearly^ so 

 that the thread is seen dividing, as it were, into two, though still 

 united by a transparent colonrless substance, of the same nature 

 as that of the septa. This may take place either equall}'- 

 throughout the whole length, or at one end first, and then 

 gradually extend to the whole, so that the difiFerent stages can 

 be seen at one time (fig. 9). After a time the secondary 

 threads pass through the same change, so as to produce a 

 band of four rows of cells ; often, indeed, before the second,aiy 

 rows are fully formed the commencement of their lateral seg- 

 mentation is discernible (fig. 10). By the continuation of 

 this process a band is formed, consisting of numerous rows of 

 cells, more or less parallel, betraying, however, the peculiarity 

 of the lyngbya-thread in the tendency to the rounding of the 

 ends, which is observed very curiously sometimes at any point 

 of injury (fig. 11). Sooner or later the process of sub-divi- 

 sion assumes the quaternary form, or its multiples, and thus 

 we have groups of 4, 8, 16, &c., indicated on the frond, by a 

 rather broader line of intercellular substance (fig. 13). Yet, 

 even in this state there is a general tendency to mark the 

 original linear origin of the frond ; though this is by no means 

 constant, the quaternary, &c., form of segmentation overcom- 

 ing it, so that the interspaces in some specimens are of uni- 

 form width throughout, and sometimes so narrow as to be 

 almost imperceptible. 



It is seldom that the quaternary segmentation occurs with 

 equal rapidity throughout the band ; but more particularly at 

 one end, whereby a fan-shaped frond is produced, which, by 

 the great rapidity and irregularity of the process, becomes 

 waved and crumpled. Should any part be arrested in growth, 

 then the other portion curls round so as to fill up the vacant 

 space, producing somewhat of a radiate disposition of the 

 cells (fig. 12). The rapid origin of a frond of 1 to 3 inches 

 in breadth, from a thread -Y^,7-^th of an inch in diameter, is 

 easily explained, Avhen it is remembered that every cell in it 

 undergoes continual segmentation in geometric ratio. 



The condition of the cell-contents varies throughout the 

 whole of these stages. As I have mentioned before in Lyng- 

 bya they are granular, possessing a few of the chlorophyle 

 utricles (Naegele) in each cell ; but as they tend towards 

 lateral segmentation they become more homogeneous, which 

 is generally continued through the whole period of growth of 

 the frond. This, however, is not necessarily so, for the con- 

 tents in a part, or throughout the whole, may be decidedly 

 granular, as is shown at fig. 15. 



After a certain time, and along one or more of the outer 



