Juty 17; 1902] 
NATURE 
275 
are wonderfully accurate. It has been shown that the 
places of stars can be measured from them with an 
accuracy almost equal to that obtainable from the 
original glass negatives. Finally, they are presumably 
permanent—far more so than the glass negatives, unless 
the toning process recently suggested by Sir William 
Crookes is adopted and found as successful as is 
expected. Against these manifest advantages is, un- 
fortunately, to be set the costliness of the process. It is 
estimated that to reproduce its 1200 plates in this way 
each observatory must have a sum of about 10,000/. at 
command, independently of the actual time spent in the 
work. This sum is large, but not prohibitive. Five 
observatories are apparently already provided with it ; 
in the interests of uniformity in a magnificent piece of 
work, may it be hoped that in some way or other the 
remaining shares will be taken up? If the paper repro- 
ductions were (as it was at one time supposed they would 
be) mere playthings of no scientific value, such expendi- 
ture might have been deprecated. But it has been 
demonstrated that they are accurate beyond expectation, 
that, in fact, an observatory provided with copies of this 
kind for the whole sky could in a few minutes obtain the 
place of any star down to the 14th magnitude with an 
accuracy equal to that with which the best meridian 
observations can be made. It seems probable that the 
outlay is as good a one as can be made with our 
present imperfect knowledge of the requirements of 
the future. 
The consideration of what this means in actual weight 
of paper brings home to us in a striking manner the 
magnitude of the whole enterprise. If the 22,000 maps 
are completed in the style adopted by the French, the 
sheets when piled one on the other would form a column 
thirty feet high and weighing nearly two tons! The 
most elaborate star atlas which has been produced up to 
the present time can be bound as a single, though rather 
large, volume, which could be added to any library with- 
out sensible disturbance. But not so with a copy of the 
Astrographic Chart ; it is a matter for the serious con- 
sideration of each fortunate possessor where and how he 
shall store the sheets and ensure their preservation. 
There is not likely, of course, to be any real difficulty in 
doing this, the point is only mentioned here to illustrate 
the novelty of the departure rendered possible by 
photography. 
As there is an obvious danger of not being able to 
carry out this vast programme (for which, it will be 
remarked, not only scientific labour, but much hard 
cash is required, and the latter may not be easy to 
extract from reluctant Governments), it is reassuring to 
know that there is at least one good alternative. We 
might carry out the work much more economically with 
a different type of instrument, though at the cost of some 
obvious advantages. The type selected in 1887, a 
refracting telescope of 114 feet focal length, allows us to 
photograph an area of the sky at one exposure limited to 
two degrees square, and 11,000 plates are required to 
cover the whole sky. Two other types were considered 
and rejected. The first was the reflecting telescope, 
with a concave mirror in place of a lens. The area 
satisfactorily photographed at one exposure with a re- 
flector is even smaller, and the number of plates required 
for the whole sky consequently greater. Though 
the reflector has distinct advantages in cheapness and in 
light-grasping power which have recommended it for 
other classes of work, there is no doubt that it was 
rightly rejected for the Astrographic Chart; all our 
experience subsequent to 1887 has tended to confirm 
this view. The third possibility open to the conference of 
1887 was the use of a doublet lens, such as is familiar in 
an ordinary camera. The lens of a camera is made up 
of two lenses (each of which is itself double) separated 
NO. 1707, VOL. 66] 
by a definite interval, where a “stop” may be inserted. 
A photograph could be taken with one of thése lenses 
alone, but only a comparatively small portion of the 
picture near the centre would be in good focus ; the com- 
bination is made to give a larger “field.” If such a 
doublet lens is used to photograph the sky, we get a 
much larger field at one exposure, and can cover the 
sky with fewer plates. The claim has recently been 
made that twenty or thirty plates would suffice to cover 
the sky instead of 11,000! Of course the results would 
be on a correspondingly smaller scale, and this extreme 
procedure is not to be contemplated as an alternative to 
the large and accurate charts with which a start has 
already been made. But if we could reduce the 10,000/. 
required to (say) 1000/., we are in the region of the 
possible or even the probable, and this only means 
reducing the number of plates required in the ratio of 
one to ten, or increasing the area covered by each in the 
same ratio. We may take it as fairly well established 
that a doublet will satisfactorily cover a field at least ten 
times as large, in area of the sky, as the single lenses at 
present in use for the work of the chart. 
The question naturally arises whether these facts were 
realised in 1887, and if so, how the single lens came to 
be preferred to the doublet. The discussion on the type 
of instrument to select took place on April 18, 1887, and 
the proces-verbaux are given on pp. 36-43 of the official 
account of the conference. Twenty-six distinguished 
astronomers were present, and eighteen of them took 
part in the debate. Zhe photographic doublet was not 
even mentioned. At the present time this circumstance 
is almost bewildering. At the end of the volume a letter 
1s printed from Prof. E. C. Pickering (who most un- 
fortunately was not able to attend the conference) 
advocating the use of the doublet, and giving detailed 
suggestions for the whole work which commend them- 
selves, in the light of subsequent experience, as admir- 
able. But his views received no attention ; the debate 
was confined almost entirely to the relative advantages 
of reflectors and refractors, and the proper size to be 
adopted for the latter, and it must be confessed that an 
opportunity was lost. Since that time Prof. Pickering, 
using doublets, has charted the whole sky himself many 
times over, while the associated observatories have not 
yet accomplished a third of their programme. It must 
not be forgotten that their programme includes much 
more than the mere charting of the sky, viz. the 
measurement of some plates and the reproduction of 
others ; but even making this allowance, the discrepancy 
between what he has done single-handed and what has 
been done on the plan preferred at Paris in 1887 is 
sufficiently serious. 
The fact is that astronomers generally were afraid of 
the doublet in 1887, and some of them have not yet lost 
their mistrust. They were afraid that so fair a promise 
was too specious; that, in fact, the gain in extent of 
field over the refractor must be accompanied by a 
corresponding loss in accuracy. At the time no definite 
information was forthcoming on this point, and it must 
be admitted that even now our knowledge is far from 
complete. It is not so easy as it might seem to test 
pictures of the stars for the minute accuracy necessary 
to an astronomer, and it may still be proved that the 
choice of the refractor in 1887 was, from the point of 
view of getting the greatest attainable accuracy, a wise 
one. But, on the other hand, it has been shown that 
the mistrust of the doublet was largely unjustifiable ; its 
accuracy is of a high, if not of the very highest, order. It 
is not even now too late to follow the excellent advice 
which was offered in 1887 only to be ignored. By adopt- 
ing the doublet the chart plates might be completed in a 
reasonable time and at a reasonable cost, though on a 
smaller scale. H. H. TURNER. 
