198 
I. ROBINSON-OBSERVATIONS ON 
valve internally. When this is completed the two portions separate, 
each possessing one old and one new valve, and thus two perfect 
plants take the place of the original one. By each of these the 
process is continued, and therefore the increase of numbers proceeds 
in a constantly augmenting proportion. The exact time occupied 
in this process of self-division is not known, hut probably under 
favourable circumstances it proceeds very rapidly. It is not, how¬ 
ever, regarded as one of true reproduction, hut is considered to he 
more analogous to that of gemmation. It would appear also that it 
cannot he continued indefinitely, hut that, after a time, probably 
from the exhaustion of the vital energy of the plant, a process of 
sporangial reproduction intervenes. The instances in which this 
latter stage has been observed are, however, comparatively rare, 
and very little is known definitely with regard to it. Professor 
Smith, in 1856, enumerated thirty-two species in which it had 
been observed, and O’Meara stated, in 1875, that twenty-eight 
others had in the interval been added to the list. The precise 
method adopted in the formation of the sporangia appears to vary 
in different species, hut the general characteristic of the process is 
the production of greatly enlarged frustules, resulting from the 
union of the cell-contents of the parent frustules. These enlarged 
frustules are doubtless the parents of a new generation, hut much 
careful observation and study are still required to trace the process 
satisfactorily throughout. One of the subsequent stages appears 
to he the production of cysts, containing many young specimens 
closely packed together, which on arriving at maturity doubt¬ 
less separate and start in life on their own account. Although 
such examples are rare, I have been so fortunate as to meet with a 
few instances in our own locality, which illustrate different stages 
of this sporangial development. Perhaps the most interesting of 
these occurred in October last, when I found in a broad ditch in 
the King’s Meads, about midway between Hertford and Ware, a 
large number of very perfect specimens of the sporangia of Cymbella 
Ehrenbergii. This species is not included in Smith’s list of observed 
species, and I am not aware that it has previously been met with. 
The sporangial frustules were much larger than their usual size, 
and much more nearly approached a symmetrical form. 
In April, 1886, I found in some material taken from the Eiver 
Beane, on Waterford Marsh, a large number of cysts of Cymato- 
pleura apiculata. Some of these contained a considerable number 
of young specimens, in various stages of development, and of 
various sizes. In the smaller specimens the wavy outline peculiar 
to this species was not apparent, although it was clearly shown in 
the larger ones. One characteristic of these cysts, which struck 
me as very peculiar, was the association of frustules of different 
genera in the same cyst. Thus I found that Cymatopleura apiculata 
was in very many instances associated with a small Surirella — 
probably 8. crumena —and, from the number of examples which I 
observed, I think the association could not have been purely acci¬ 
dental. I also found a number of similar specimens amongst some 
