246 
PROFESSOR FORBES ON THE EXTINCTION OF THE SOLAR RAYS 
54. The mean of the whole, giving- to each series its proper weight, is 1T54; but 
considering the variations of the second, third, and fourth series, and the extremely 
favourable circumstances and good agreement of the fifth, I prefer to adopt it alone, 
and assume for the factor of reduction G. 7- to B. 2 1-168*. 
Reduction to Intervals of One Minute. 
55. In the instructions which Sir J. Herschel had provided for me, it was observed 
that the velocity of heating or cooling might be noted for thirty seconds instead 
of one minute, and reduced to the standard unit by doubling it. But this does not 
appear to be exact, and in order to compare observations made with thirty-second 
intervals with those made at sixty-second intervals, a greater factor than 2 appears 
to be necessary, for reasons not difficult to anticipate. Whilst therefore I agree with 
the later instructions in preferring sixty-second intervals, it is useful to have a factor 
for reduction'!'. 
56. Professor Kamtz, by careful and multiplied observations on the actinometer B. 2. 
at the Faulhorn, found the rise in 15, 30, and 60 seconds, to be proportionally 1, 2345, 
5 - 208 in the MS. notes with which he supplied me, and which almost exactly coincides 
with what he has stated in his work on Meteorology, vol. iii. p. 21. Hence the factor 
of reduction of thirty-second to sixty-second intervals is = 2-224, which I have 
employed when necessary. 
57- The following Tables contain the meteorological observations on the 25th of 
September 1832, at Brientz and the Faulhorn, with the reductions necessary to render 
the results immediately applicable. 
58. The observed times were nearly mean time at the place. They are reduced to 
apparent time , which is used in all the calculations and projections in which the acti- 
nometer is employed, on account of the facility which it affords for the direct com- 
parison of observations at equal altitudes before and after noon. 
59. The barometers used were both on the syphon construction : that at the lower 
station was divided on the old Swiss plan of French inches, lines, and 16ths with 
double readings. The upper barometer was divided into French lines and decimals, 
and was reduced to zero, Reaumur, by M. Kamtz, before communicating the read- 
ings to me. I have reduced the other to the same temperature by means of its at- 
tached thermometer. Both barometers have been reduced into millimetres, which 
has been assumed as the standard of calculation (and 760 millimetres as the mean 
atmospheric pressure) for reasons W'hich I need not now specify. The barometers 
were compared on the 24th of September and found to agree. Their index errors, as 
* The observations at Geneva (first series), which are the best of the others, show that there is no reason 
for believing that the instruments had changed in any way at the date of the fourth series. 
t Where by accident a single observation has been extended to seventy or seventy-five seconds, a simple pro- 
portional reduction will be sufficient, as shown in one of the examples already quoted. 
