612 
PROFESSOR ROSCOE ON A METHOD OF METEOROLOGICAL 
I. 
II. 
III. 
I. 
II. 
HI. 
20 
2-30 
0-10 
70 
1-47 
0-022 
30 
2*10 
0*09 
80 
1-28 
0-010 
40 
1-90 
0-02 
90 
1-07 
0-045 
50 
1*76 
0-016 
100 
0-916 
0-053 
60 
1-62 
0-013 
110 
0-830 
0-056 
120 
0-755 
0-050 
A comparison of the several curves of the graduation of strip A found in Plate XXVIII. 
fig. 1 shows that the determinations agree as well as can be expected from such photo- 
metric experiments ; the mean tabular error between the positions 40 and 80 millims. 
on the strip not exceeding one per cent, of the measured intensity. 
For the second method of graduation sheets of paper tinted by lithography of a 
brownish colour and of different shades are employed, and a portion of each sheet is cut 
out, so that the several tints differ considerably from each other, and correspond to the 
tints taken at definite intervals along the fixed strip. These are then gummed over half 
the reading block, and the value of each read off on several pendulum strips, the inten- 
sity of which had previously been determined by the normal tint. Having thus obtained 
the intensity of each of the fixed tints, the fixed strip is graduated in terms of the pen- 
dulum strip by determining the points on the former equal in intensity to the fixed tints. 
This method possesses several advantages over that just described, and is to be preferred 
to it, although the comparison is an indirect one, as the intensity of the fixed tints can 
be found with a great degree of accuracy by repeated measurements ; and when their 
intensities a.re once determined they can be preserved for a length of time, as they do 
not undergo any change of shade, and therefore can serve for the graduation of a large 
number of fixed strips ; the preparation of which is accordingly not dependent, as is the 
case in the first method, upon the state of the weather. 
The following numbers may serve as an example of this method : 
(1) Determination of the intensity of fixed tints upon pendulum strips. 
No. 1. 
No. 2. 
No. 3. 
No. 4. 
No. 5. 
No. 6. 
No. 7. 
No. 8. 
No. 9. 
No. 10. j 
No. 11. 
No. 12. 
Reading of normal tint on pendulum 1 
strip J 
Reading on pendulum strip of fixed 1 
tint No. I. J 
„ No. II. 
„ No. III. 
„ No. IV. 
„ No. V. 
153-2 
mm. 
82-0 
mm. 
131-1 
mm. 
136-6 
mm. 
1057 
mm. 
17-5 
mm. 
121-6 
19-2 
mm. 
51-9 
mm. 
131-6 
mm. 
119-4 
mm. 
98-0 
40-3 
90-4 
115-7 
297 
1087 
50 
50-2 
91 0 
159-5 
67-2 
1007 
25-4 
661 
50-5 
125-8 
990 
s ’ i -2 
120-6 
93-0 
1500 
157 
50-1 
24-5 
52-3 
89-3 
145-5 
22-6 
517 
134-0 
The intensities for each determination of a fixed tint are obtained from the above 
numbers by dividing the numbers found in Column II. of Table I. (p. 607) opposite the 
millimetre readings of each fixed tint by those found in the same Table opposite to the 
readings of the normal tint. 
