42 Dr Greville’s Descriptions of New Genera 
valve might serve to show more exactly their relations ; but 
the probability of obtaining a side view of the frustule is so 
remote, that I prefer giving figures of these interesting 
objects at once, to laying them aside for an indefinite 
period. It only requires to be understood that their posi- 
tion in the genus is provisional. There is no reason why 
the same latitude should not be allowed in the publication 
of Diatoms as in that of other plants; and especially of 
other alge. Forexample, while engaged in the description 
of these Amphiprore, the 49th Part of my friend Professor 
Harvey’s admirable “Phycologia Australica” has appeared,in 
which, under Cryptonemia dectpiens, he remarks, “ In now 
referring it to Cryptonemia, I must be understood to do so 
provisionally, until the discovery of its fruit enable us to 
assign it its proper place in the system.” ‘This is a most 
reasonable privilege. It may be observed, with regard 
to the frustules now before me, that there is all the ap- 
pearance of a distinct wing, which becomes narrower as 
it approaches the constriction, as in the known Amphr- 
prore. 
Amphiprora? undulata, n. sp., Grev.—Small; front view 
quadrangular, with rounded angles, and nearly straight 
sides, and a slight, somewhat notch-like constriction ; sur- 
face rather coarsely striated, and undulated with longitudi- 
nal ridges, the two principal ones curving outwards, and 
passing just within the constriction. Length about -0030" 
(fig. 22). 
Hab.—Curteis Straits; in a dredging communicated by 
Dr Roberts. 
This species is as rectangular in its outline as A. pusilla 
of Gregory. The two outer prominent ridges or folds evi- 
dently occupy the place of supplementary wings, while the 
inner ones are merely undulations, sufficiently decided, 
however, to throw the striz which cross them out of focus. 
I have examined a number of examples, but have not 
succeeded in finding a valve. The latter would no doubt 
assist in determining the genus in the case of aberrant 
forms; but I apprehend that in most instances the specific 
character will be best obtained from the front view. 
