The Development, &c., of Diatomacer. By Dr.G.C. Wallich. 77 
the two principal spines are T-shaped, the base of the T being 
attached to the valve; and the tubule (which may be very dis- 
tinctly seen in the larger specimens) having its external opening 
immediately at the top of the stem of the T, and not bifurcating 
so as to open at the two extremities of the cross-bar- of the T, as 
would be the case in any sponge spicule. There is strong evidence 
in support of the view that, through these tubules, a communication 
is established either with the surrounding medium, or, in some 
filamentous genera, with the corresponding spines of adjacent 
frustules. Indeed, in Creswellia the extraordinary fact presents 
itself that every one of the large series of spines forming the 
coronary circlets is precisely in apposition with a similar number 
of spines in the adjacent frustules. And from the circumstance of 
the spines not being tapered but cylindrical, and, if anything, 
slightly inflated at the points of union with the adjacent spines, 
they very closely resemble in appearance, and are in all probability 
identical in function with, the horny-looking processes met with in 
some of the filamentous Desmidiacex.* I mention these morpho- 
logical resemblances because the more I look into the structural 
details of the Desmidiacex and Diatomacex, the more closely do I 
find that they agree on all points apart from the degree of con- 
solidation, and the nature of the substance producing the consoli- 
dation, of their enveloping wall. By far the most important proof, 
however, of the close relation between these two families has yet to 
be submitted, when I come to treat of the “ sporangial frustule.” 
Before passing on to the next section of my subject, I have to 
draw attention to an observation of Dr. Macdonald’s, which cer- 
tainly appears to me to involve more than he intended—an obser- 
vation which, if correct, would form an immovable keystone to his 
hypothesis; but if incorrect, as I believe it to be, cannot fail 
seriously to mislead. He says (loc. cit. p. 3), “ It stands to reason 
that as the two new half-frustules are developed endogenously, or 
within their parent, the former must be smaller than the latter by 
the whole thickness of the siliceous investment,” &e. &e. And 
again (p. 4), “ Yet it is actually within the fully formed valve that 
the new half-frustule is produced; and if so, it must, as before 
stated, be smaller than its parent,” &e. &e. 
I venture to affirm that the new half-frustule is never developed 
within the fully formed or parent valve at all; and that it is 
invariably developed as an outgrowth from the soft contents, 
entirely beyond the free margin and plane of the parent half- 
frustule; precisely as the new segment of the Desmid is developed 
* The close resemblance here referred to may be well seen on comparing the 
figures of Creswellia, on the one hand, and Streptonema on the other, appended to 
Dr. Greville’s “ Description of New and Rare Diatoms” (‘ Q. J. M.S. Jan. 1864), 
and my papers on “ The Desmidiacew of Bengal” (‘ Annals and Mag. Nat. Hist.’ 
ser. 3, vol. v. plate viii. fig. 1). 
a 2 
