340 
POPULAR SCIENCE REVIEW. 
he gives without question, nay, with implicit sanction, a series 
of 44 heterogenetic ” phenomena from 64 the much neglected 
memoir of Dr. Grros” and others. The former has been for 
twenty years before the scientific world, during which time un- 
paralleled vigour has- been displayed by biologists in every de- 
partment of research, and especially in the development of 
minute life-forms ; which have been studied with constantly 
improving apparatus. But during the whole of this time not a 
single instance of accepted corroboration of these strange trans- 
formations can be pointed out. But thousands of observations 
have been made and recorded that are directly adverse to the 
whole. Surely this alone should have suggested caution, and 
the careful and competent repetition of both Grros’ and his own 
observations. But it was not so. 
All who have any practical knowledge of the nature of such I 
enquiries will admit that an indispensable condition — a sine qua 
non — to accurate results, is continuity of observation ; and that, 
moreover, upon the same organisms, and with the very best ap- 
pliances ; all of which should he guarded by the elimination of 
every conceivable source of fallacy. 
Discontinuous or interrupted observation is, in such inquiries, - 
worse than useless. It is at once a prolific and a fascinating 
source of error. . The same must he said of not working out the 1 
history of the same individual. In all such researches scores, • 
nay, hundreds of hours are wasted — or rather apparently lost — 
in following the organism to a certain point, and then it dies ; 
or some accident happens, or some distraction to the observer 
arises ; and the temptation is to take up the observation again 
upon another form in apparently the same developmental 
condition as the former one was in when the interruption 
happened. No source of error can he more serious in practice, 'J. 
when the objects of research are so minute and unknown. Hun- 
dreds of instances of this might be given. Indeed there can be ! 
no accuracy unless the observation begins again ab initio , and 
is carried persistently to the end. 
Let any one take a moderately decomposed infusion of fish 
or brain, and put a small clear drop of it from the point of a 
fine 44 dipper into half a wine-glass of Cohn’s nutritive fluid, and ; 
leave it for a few hours. Let a drop of this be put on in the i 
usual way upon the continuous moist stage described in the l 
44 Researches into the Life-History of the Monads,” and let | 
a bi. objective he used. The probability is that for two 
days nothing will be visible to the most careful scrutiny but 
bacteria — at least it can be so arranged. But now, the observer 1 
who spends two or three hours discontinuously at the instrument 
will probably observe that what look like some of the bacteria 
are getting larger ; and if then a night should intervene, he will. 
