368 MORTEN P. PORSILD. 
In Table III I have compared the numerical mean from Disco of 
our observations as well as the two Danish series published by 
STEENSTRUP, the latter having been corrected to the absorption- 
coefficient of the paper used by me. As shown by the Table, and 
as has already been pointed out by STEENSTRUP, the fall and the 
rise of the monthly means is very regular, 1.е. during longer periods 
of the year the light-totals are more dependant upon the 
altitude of the sun than upon cloud-screens. 
The dependence of the daily totals of the actinic light upon 
the altitude of the sun is still more distinctly shown when we con- 
sider the mean number of the periods which have about similar 
positions of sun and lengths of days. This is shown in Table IV. 
There the year is divided into 24 periods of different lengths, and 
these periods are arranged according to the declination of the 
sun. The Table shows how regularly the mean totals increase 
and decrease without inversions. Besides the actinometer num- 
bers the Table contains in the two lower lines the relative 
values of the numbers in question. As the unit for these relative 
values, the approximately annual mean numbers from South Disco 
(48) have been chosen and the value in Table II has been deducted 
from it. The values should consequently be understood as fol- 
lows: — In the Disco-series the light-total rises to 25 
times as much as the annual mean number, and sinks to 
1155. In the Anholt-series it rises to twenty fold, and 
sinks to Из. The annual mean number for Anholt (56) is 
about three times as high as that of South Disco. 
The Actinic Light-totals during lower positions of the Sun. 
On considering the mean light-totals of the winter in Table IV 
we see that they correspond very closely with the actinometer num- 
bers during low positions of the sun, more particularly closely in 
the Disco-series. This means that when the sun stands near 
the horizon or almost below it the cloud-screen plays 
almost no part at all. The differences in the totals for one 
period and for another are great, the inversions are not done away 
with and the totals fall and rise symmetrically around the minimum 
of the sun’s altitude. Conversely, the cloud-screen plays a great 
part during the high positions of the sun. The differences from the 
one period to the other become small and the inversions frequent. 
These facts may be explained from the well-known absorptive 
powers of the most refrangible rays in the atmosphere. A. PAULSEN, 
in "Ugeskrift for Læger” (Kbh., 1903), has given the following Table 
