PROCEEDINGS OF SOCIETIES. 221 
something like a starting point in the original Navicula rhomboides 
—they had its shape, its markings, and its median line, it was figured 
in “Smith,” and would be a very good one to start from. Then in 
the same book they had also N. crassinervis, or crassinervia, which had 
distinctly different ends, was round at the sides, and had the ends of 
the median line terminating in distinct nodules, compared with which 
N. rhomboides had its sides straighter, was more generally rhomboidal, 
and its median line terminated in blunted lancet-shaped points. 
Between these two they could easily get every variation leading 
from one to the other, and he thought it was most important from 
amongst these varieties to get a well-established typical Rhomboides 
and a typical Crassinervis. What had happened was that for some 
time they had one form only, after that a coarser form was brought 
forward as a test for a+ inch; it was much rounder at the sides, and 
there was a difference in the centre and ends of its median line, 
which were flattened and battledore-shaped; and then after this 
they got another form from the Cherryfield diatoms. On the same 
slide they might see the original Rhomboides and the new specimen. 
He thought that in this respect Moller had done a great deal towards 
creating this confusion, for on his type slide they found Rhomboides; 
but it was the big form, and nothing at all like the Rhomboides of 
Smith, whilst next to it was what he called Crassinervis—exactly 
the same as the original Rhomboides, but nothing at all like the 
Crassinervia of Smith. He should be one of the last to increase the 
number of genera and species upon insufficient data, but thought they 
ought not to group together forms which differed so greatly from 
the type given as that of the species. They wanted at least to know 
what they were talking about, instead of getting one so mixed up 
with another that as test-objects they might be used for what might 
be called illegitimate purposes and so became of little value. 
Mr. Slack said the paper raised a very wide practical question, 
and one which was not confined to diatoms. When a botanist was 
able to show that certain extreme forms were connected by a series of 
intermediate gradations, he was justified in placing them together in the 
same group notwithstanding their differences, just as a greyhound and 
a bulldog were both classed as dogs. In order to know what they 
were talking about, where there were a great number of so-called 
species in the question, it was well to retain distinctive names for 
them for purposes of identity even long after it had been shown that 
they had no claim to be spoken of as belonging to different genera or 
species. 
Mr. Charles Brooke said it appeared to him as a general question 
that it was quite impossible to assign two individuals to different 
genera when there was a series of intermediate forms which might be 
found passing from one to the other. It might be desirable to know 
the various forms by different names—as in the case of dogs already 
mentioned—but they should not be known as distinct species, and this 
should never be done where such gradations existed. This gradation 
of form had been traced throughout the whole class of the Foramini- 
fera to a far greater extent than amongst the diatoms, and amongst 
R 2 
