492 Analyses of Books, [August, 
all its primary assumptions, would be wilfully to foreclose the 
possibility of ever making way in the science of numbers.” 
A further contention, and one of very grave importance, is that 
“ not only in beginning a course of physical inquiry, but in every 
stage of it we have to walk, by faith quite as much as by sight.” 
In support of this declaration, a number of striking instances are 
given. “ A philosophy restricted to one’s own experience could 
be nothing better than a volume of travels by Simon Stylites.” 
Even Darwin could not dispense with the observations and ex- 
periments of others, and was in consequence beholden to testi- 
mony. 
But here comes an important consideration which the author 
has overlooked, and which, as far as we can call to mind, has 
not met elsewhere with due attention. Testimony in Science, 
and testimony, e.g., in history, do not by any means stand upon 
the same footing. On the contrary, they differ in an important, 
if not the most important, respeCt. This will appear from the 
following illustration. We read, say in the Coniptes Rendus , that 
Thalen has examined a newly discovered metal, and has found 
in its specftrum such and such characteristic lines. Why do we 
accept his testimony? Not merely because we know him as a 
skilful, thoroughly-competent speCtroscopist, and as a man of 
unquestioned integrity who could, moreover, have no intelligible 
motive for falsifying or suppressing his observations. We 
receive his account of the speCtrum because it is open to others 
to test and check his results. We know that others will have 
already done so, and we are well aware that Thalen himself has 
worked under the full expectation of such a revision. Hence 
he will have been equally on his guard against deceiving himself 
or attempting to deceive others. 
With testimony in history this is not the case. We cannot 
send anyone up the stream of time to verify the statements of 
old chroniclers, or of monumental inscriptions and hieroglyphics. 
Testimony in Science is like a guide who, showing us round some 
palace, should throw open door after door, and whilst telling us 
the contents of each room, should give us the opportunity to 
examine for ourselves. 
Historical testimony, on the other hand, is like a guide who, 
whilst describing the various rooms, should bid us take his word, 
as the keys were lost. 
One postulate which must be granted, in faith as a preliminary 
to our scientific studies, the author does not bring into full relief. 
We refer to the necessity of assuming, without the possibility 
of ever proving, that the impressions made upon our senses are 
accurate or at least comparable pictures of an objective world. 
We cannot, however, follow Mr. Griffith any further. Dissent- 
ing as we must from the Teleology which makes so prominent a 
figure in his fifth and sixth chapters, and which we believe merely 
strengthens the hands of the school of thinkers which he is 
