On the Classification of the Diatomaceae. By M. Paul Petit. 75 
The frustules are discoid (? spherical), elliptical or cylindrical, and 
are united in pairs or in greater numbers, forming filaments more 
or less lengthened, thus giving them a pseudo-likeness to Algae of 
the family of the Confervacese. 
Genera, Cyclotella, Melosira. 
These are the affinities of the tribes, as I understand them. I 
may add, in concluding the above arrangement of the Diatomaceae, 
that it is only a very incomplete essay. However, I shall esteem 
myself happy if the criticisms of which the system I now pro- 
pose may become the subject, should be the means of dissipating the 
obscurity which has so long enveloped the physiological knowledge 
of the Diatomaceae. 
[It affords me much pleasure in introducing the labours of my 
friend M. Petit to the readers of this Journal ; and although, as 
he remarks, the essay is incomplete, it will have served a very 
useful purpose if it induces the diatomist to study the vital portion 
as well as the “ dry bones ” of the Diatomaceae. One very 
important phase in the life - history of these organisms might 
perhaps be discovered by patient and continuous observation of the 
living frustule, viz. the formation of microspores. Their repro- 
duction by means of auxospores and multiplication by self-division 
has long been known, but the existence of germs has not yet been 
detected. Their discovery would help to explain the rapid appear- 
ance of diatoms in recent rain puddles or amongst the confervoid 
growths invariably seen on walls and paths in the neighbourhood 
of leaky rain-water tanks and butts. It would be of the greatest 
interest to detect if the plasma in the frustule retains its vitality 
after the frustule becomes dry and itself an impalpable powder, and 
in this state carried hither and thither by the wind, and when 
deposited in a suitable nidus germinating. 
Dr. Pfitzer quotes with approval Meneghini’s observation, 
“ That anatomy has to effect the same useful revolution in the 
natural classification of diatoms as has been produced in the system 
and nomenclature of the Conchylia.” Unfortunately there is but 
little for the diatomist to anatomize in the organisms he studies, 
and that little is so rapidly altered by death, that it appears to me 
impossible to classify them in accordance with the disposition of 
the endochrome ; we can in a vast number of instances only infer 
this disposition from the resemblance the dead frustules and valves 
bear to those which we have been able to study whilst living, thus 
constructing a system partly natural and partly artificial. 
Professor H. L. Smith remarks,* “ As for Dr. Pfitzer ’s classifica- 
tion, I only know it through abstracts published in the journals : that 
method I long ago tried and abandoned, as he will be obliged to do. 
With all the enthusiasm of a novice, I fancied here was the key to 
* ‘ M. M. J.,’ vol. ix. p. 221. 
VOL. XVIII. 
G 
