262 
POPULAE SCIENCE EEVIEW. 
bility of dispensing with the observer has been fully shown at Paris, for some 
years past, by the experiments made by M. Poiro, at my suggestion, in his 
ateliers, with the assistance of MM. H. Eobert, Digney Preres, and Jumet. 
“ The process, which is extremely simple when operating with the sim, 
becomes more dehcate, but not impracticable, when applied to the stars. It 
consists in substituting a photogTaphic ^date for the eye of the observer, and 
to register electrically the instant when the light is admitted into the camera 
applied to the meridian telescope.” We have thus obtained ten observations 
of the sun in twenty seconds. When I say we have obtained, it would be 
more correct to say that we thought of making an impromptu astronomer of a 
boy, who was simply charged with drawing a shutter and moving a spring, a 
task which might have been performed by a machine. If astronomers who 
have already separately adopted these two powerful means of observation, con- 
clude, as I hope they will, by adopting the still more powerful combination, I 
desire that the Academie may be reminded that I had the honour of present- 
ing to it, some five or six years ago, the first effectual observation of an 
eclipse completely registered by means of this combination of electricity and 
photography, and the first meridian observation of the sun, independent of 
the senses of the human brain."^ 
“ While examining recently the negative of this curious observation, which 
I have carefully preserved, I remarked certaui defects which I had not 
noticed before, and upon examining it more closely, I recognized that these 
defects were inherent, not in the proof, but in the sun itself They are, in 
fact, solar spots which have described themselves at the same time as the 
edges* of the disk to which the attention of the observer should be directed. 
Here, then, is an automatic observation which gives with great exactness, not 
only the situation of the orb at the moment of the true southing, but also 
that of its spots, the study of which has assumed so great an importance of 
late years. 
“ Here, then, is a fresh example of the superiority of automatic observation 
over the old method, based upon our senses ; as a general rule, we see only 
those things which interest us at the moment of observation, and the rest 
almost always escapes pre-engaged attention. Automatic observation, on 
the contrary, registers everything which we seek now and which we look for 
afterwards. 
“ It has been objected that the conij)lication this system will introduce into 
the daily practice of our observatories will provoke much opposition ; to this 
I reply, that we must pay the same price for every increase of precision in the 
celestial measurements ; for an observatory of the nineteenth century differs 
even more from the early observatories, than the machinery of our mills 
differs from the old spinning-wheel. We shall make fewer observations, but 
these few will be more reliable. A recent instance has shown to astronomers 
how unsafe it is to boast of apparent exactness. It is well established now- 
a-days that the distance of the earth from the sun, accepted till recently by 
all astronomers as a definite datum, included an error of more than one- 
thirtieth of its value. We must expect other mistakes of this kind so long 
“ By this I mean negatives obtained upon a large scale, with the indica- 
tion necessary to refer the measure to the celestial circles.” 
