875 .
] Gw the Lines of the Solar Spectrum .
201 Adopting the above numbers as the respective numerical values of each rder , we have for barracks:\#151 ; No. of Value No. of order .
observations .
of order .
Total .
2 .
89 X 2*13 = 189*57 3 88 X 3*46 = 304*48 4 \amp ; 5 97 X 4*66 as 452*02 Sums. .
, ... 274 946*07 giving a mean of 3*45 .
For hospitals we have 2 20 X 2*13 = 42*60 3 46 X 3*46 = 159*16 4 \amp ; 5 20 X 4*66 = 93*20 Sums. .
... 86 294*96 giving a mean of 3*43 .
Here we find the same numerical value ( signifying close ) applied to 0*580 in hospitals and 0*601 in barracks .
There is thus , even in this comparatively limited number of observations , a confirmation of the opinion that more air is necessary to keep an air-space sweet in disease than in health .
It is , however , right to point out that in the one case the occupation was continuous , and in the other chiefly at night only .
II .
" On the Atmospheric Lines of the Solar Spectrum , illustrated by a Map drawn on the same Scale as that adopted by Kirch-hoff .
" By J. B. N. Hennessey , F.R.A.S. Communicated by Prof. Stokes , Sec. R.S. Received , the map June 9 , 1874 , the text January 11 , 1S75 .
( Abstract .
) The spectroscopic observations described in this paper were made with instruments belonging to the Royal Society , and in accordance with certain suggestions which had been made to the author by a committee appointed in consequence of a letter of his to Sir Edward Sabine , President , dated 13th February , 1866 .
In view of his residence at a considerable height above the sea-level , and of the exceedingly clear atmosphere prevailing at some periods of the year , it was suggested that the locality was peculiarly favourable for a determination of the lines 0 the solar spectrum due to atmospheric absorption , and that , for this purpose , the solar spectrum when the sun was high should be compared with 202 On the Lines of the Solar Spectrum .
[ Jan. 28 .
the spectrum at sunset , and any additional lines which might appear in the latter case should be noted with reference to Kirchhoff 's map .
Accordingly the author set to work with the spectroscope first supplied to him , and in the autumns of 1868 and 1869 mapped the differences in question from the extreme red to D. These results appeared in the ' Proceedings of the Royal Society * for June 16,1870 , and the map of the spectra , sun high and sun low , of the region in question forms plate 1 of the 19th yolume .
The instrument first* supplied to the author was found in practice to be of insufficient power to permit of ready identification of the lines seen in the spectrum of the sun when high with those represented in Kirchhoff 's map ; and a new spectroscope of greater power was supplied to him , which reached him at the end of the year 1871 .
Observations for a continuation of his map had , in the mean time , been taken with the old instrument in the autumns of 1870 and 1871 , and the spectrum mapped from D to F , in continuation of the former map .
But the new instrument proved so superior to the old , that the author determined to map the whole spectrum afresh from observations made with it , using the former maps merely as skeleton forms .
The observations with the new instrument were carried on in the autumns of 1872 and 1873 , and the map now presented is the result .
Observations were also made to ascertain whether any of the lines which came out when the sun is low , especially those which are also seen , but narrower and less conspicuous , when the sun is high , could be due , not to specific atmospheric absorption , but to the general weakening of the light , causing parts of the spectrum already weakened by solar absorption to appear dark when a general weakening of the fight was superinduced , though they had appeared bright when the fight was strong .
For this purpose the spectrum of the sun when high , as seen in the usual way , was compared with the spectrum when the intensity was artificially reduced in various ways .
The best comparison was obtained by taking advantage of a natural phenomenon .
At Mussoorie , late in the autumn , a haze , visible at sunset , extends over the low country , and grows day by day in height , till it causes the sun virtually to set in haze while still 3 ' or more above the horizon , whereas in the clear season it is visible till it attains a depression of 11 ' .
The result of the comparison was , that none of the additional fines were discovered have any other origin than selective atmospheric absorption .

