On “ Navicula erassinervis,’ &e. By W. H. Dallinger. 5 
question are but conditions of the one form (Ahombozdes), it appears 
to me that some interest attaches to the fact that proofs of this 
may be obtained by a serial study of the silicious frustules, from 
the smallest or most dwarfed, to the largest or most developed ; and 
that such a study brings out plainly what value should be attached 
to this diatom as a test-object. 
I have now a very considerable collection of these forms, 
gathered from many sources, and through the courtesy of many 
friends; but I rarely find, amongst the very finest and minutest of 
them, any frustules that will not yield complete resolution into 
hemispheres, with Powell and Lealand’s “new formula” }-inch 
objective (immersion), with one or other of the methods of illu- 
mination I now employ; using as the source of light a broad- 
wicked paraffin lamp with the flame turned edgeways to the con- 
denser. Unfortunately time has never been sufficiently at my 
disposal, to enable me to make micro-photography my servant in 
study. But as the objects I have concerned myself with during 
the greater part of my life as a microscopist, have been swiftly 
moving vital ones, in the portrayal of which even photography 
could have been of no service, I have been obliged for some years 
to cultivate, with as much care as I could exercise, an accurate and 
delicate use of the pencil, in fixing permanently the images of 
objects which the microscope revealed. I believe in this instance 
its aid may be called in with some service. 
Having satisfied myself of the absolute similarity, in ultimate 
structure, of all the modifications of what we shall hereafter know 
as N. rhomboides, it appeared to me advisable to take one of the 
smallest of the forms hitherto labelled “ N. crassinervis,” and find 
the least magnifying power that would perfectly resolve it, and 
then employ the same power, and the same mode of illumination, 
on a series of the increasingly larger forms, until the extremely 
coarse ones were reached. In this way the identity of the ultimate 
structure in every developmental condition would be seen. 
The power necessary to resolve completely the smaller forms of 
the “ N. crassinervis” in my cabinet was 800 diameters (obtained 
with the }th above referred to). The figure marked A, Plate CLXV., 
represents the result. The brilliant definition obtained by the lens, 
rendered much more apparent the “resolution” than the most 
delicate drawing or photograph could do; but the drawing given 
is a careful rendering of the effect. It will be seen that some 
portion of the striae (transverse) are seen on the left-hand side of 
the midrib: but they are seen perfectly on the right-hand side; 
and in the upper half of it, resolution into dots, or hemispheres, 
placed in rectangular rows, is, in a good light, clearly visible. 
Of course, it is not pretended that enumeration of striz, and 
so forth, could be effected from this or the following drawings, as 
