104 REPORTS ON THE STATE OF SCIENCE. 
deduced from the results for the thirteen stations and for Strassburg and 
Munich separately :-— 
Annual Variation in H.. 
aa | Jan. | Feb. | Mar. | April | May | June | July | Aug. | Sept. | Oct. | Nov. | Dec 
Mean of 13 Stations | 103 -10°4 91 101 | 105 | 10°77 | 10°99 | 11°4| 10-4] 11:9) 10°8 | 10-1 
Number of Cases .| 26 | 22 32 a9! "| “81 27 24 61 46 38 25 25 
Munich . = -| 10°0 | 10°4 OF | Bed) LD | ar) | DEAF 120! 1053) Less 118 | 114 
Number of Cases . 4 | 3 6 4 | 2 4 3 11 5 een ae 5 
Strassburg . .| 10° | 10°6 94 94 / 106 | 10:9 | 10:8) 12:3) 10:9] 11°9 |} 11:0] 111 
Number of Cases . 5 | 5 BW le & 5 4 9 | 8 6 | 6 5 
| 
Annual Variation in T.. 
den eed SL ——— — 4 eS ae 
| ——- Jan. | Feb. | Mar. | April| May June | July | Aug. | Sept. | Oct. | Noy. | Dec. | 
Mean of 13Stations| 13 | 11 | 16 | 16 | 17 | 20 | 20 | 18 | 92 | 14 | 45 14 
Munich . 5 | 14 10°") 46 19 25 20 15 16 26 9 10 13 
Strassburg . Salt eh 10 | 16 20 17 17 21 15 | 23 13 12 10 | 
The values of H, are plotted in Fig. 1. 
The observations are made at the beginning of each month, and the 
results may be taken to correspond nearly to the 4th of each month. The 
observations made at the end of July are counted in August, so that the 
values for August correspond nearly to the last day in July. The most 
remarkable feature is the depression in the value of H, i March and 
September. 
The fact that H, is greatest in October and T, is greatest in September 
effectually refutes the hypothesis that the cessation in the fall of tem- 
perature is due to any effect either of solar or of terrestrial radiation on 
the instruments. 
The value of T, is certainly slightly greater during the summer months, 
but the difference between the mean value for summer and winter is only 
6° C. This difference is less than the difference in the mean temperature 
of the atmosphere up to 10 km. for the corresponding period, and it may 
well be a real increase in the temperature of the advective region owing to 
increased radiation from below. 
It has been suggested that the cause of the sudden change in the 
temperature gradient, which would naturally be expected to diminish 
gradually, is the formation of a veil of Ci. or Ci. S. If this were the case 
the annual variation in the height of these clouds ought to show the same 
peculiarities as the annual variation in H,. The annual variation in the 
height of Ci., Ci. S., Ci. Cu. is given in Fig. 1. The observations used 
are those for the international year 1896-97. The curves do not show 
the very marked minima in March and September which occur in the 
H, diagram, but there are indications of a peculiarity of this kind, more 
pronounced in the Ci. 8. and Ci. Cu. curves than in that for Ci. proper. The 
annual range is about the same as for H,, 2°5 km. nearly. The actual 
values of the heights are, however, much less than those of H,. Thus, 
while the results point to some common cause for the variation in H, 
and in the height of the clouds, they indicate that the formation of clouds 
is not a usual cause of the sudden fall and change of sign in the temperature 
gradient. 
One of the most remarkable features is the large variation sometimes 
found in T, and H, from one day to the next. This has been most marked 
in England, and the general agreement in results from different stations 
