50 Address of the President at the Annual Meeting. 
species-making, which is one of the greatest obstacles to all 
truly scientific progress, alike in systematic Zoology and 
Botany, in the Geography of plants and animals, and in 
Palaeontology. I am happy to say that in regard to the 
particular subject which suggested these remarks, namely, 
the range of variation among the Diatomacece, I find the 
highest British authority upon that tribe, my friend Professor 
W. Smith, to be quite in accordance with me ; and strongly 
recommend an attentive perusal of his paper ' On the deter- 
mination of Species in the Diatomaceae' (' Microscopical 
Journal,' January 1855) to all who are engaged in the study 
of that group. I would particularly direct their attention to 
his remark, that to adopt as specific a shape or size that may 
occur in all the individuals, however numerous, of any one 
gathering, is liable to lead to fallacious results, from the 
circumstance that all these specimens may have been pro- 
duced by self-division from a single individual ; and that as, 
in each such division, the characters of the original are 
perpetuated, a modification of the outline of each individual 
in the entire collection, may arise from some accidental cir- 
cumstance which affected that of the primordial frustule. 
Hence neither size nor outline are sufficient to mark the spe- 
cies among Diatomaceae : it is requisite to compare specimens 
from different localities, in order to determine with tolerable 
certainty, what is the average or typical form and magnitude 
of any species ; and its range of variation must be determined 
by a comparison of forms brought from every known habitat^ 
before the extent of departure which should be held to consti- 
tute a distinct species, can be even approximately determined. 
Thus Professor Gregory shows us (loc. cit.) that under the 
general category of Navicula varians may be ranked a consi- 
derable number of forms, differing widely in shape, but seem- 
ingly tolerably constant in number of striae (the range of 
variation of these being apparently from 14 to 18 in '001", 
and the usual number being 16), which have been described 
under different specific names ; whilst there is another very 
analogous group of reputed species, but also seeming to be 
tolerably constant to a very different number of striae (their 
range of variation being apparently from 24 lo 26 in '001"), 
which he proposes to designate as N. mutahilis. Now it must 
be considered an open question, to be only decided by the 
examination of collections from a much greater number of 
localities, whether these two comprehensive species are not 
themselves in reality but one ; since it is quite certain that 
there are species in which the range of variation in the 
number of striae is much greater than from 18^ the greatest 
