338 REvORT—1905. 
6. Note on Professor Kapteyn’s Method of Determining Stellar Parallax 
by Means of Photography. By Artour A. Rampaut, M.A., D.Sc., 
F.R.S. 
Professor Kapteyn’s method is, described in Part I. of the publications of the 
Astronomical Laboratory of Gréningen. Instead of exposing the same plate at or 
near three successive epochs of maximum parallactic displacement, as Kapteyn 
requires, it is proposed in this paper to take three different negatives, which may 
be developed in the ordinary course as soon as convenient after exposure. From 
each of these negatives a positive is printed, and these three positives are trans- 
ferred again to a single final negative. 
In order to determine the probable error introduced by this double photo- 
graphic reproduction, two plates were compared, each containing a triple photograph 
of a group of stars. In one case the plate was exposed in the telescope three times, 
in rapid succession, to the same region of the sky, the telescope being slightly dis- 
placed between each exposure. In the other a negative was taken, and from this 
three separate positives were made, which were afterwards transferred to a single 
plate. The probable error of a single measure of the distance between two images 
of the same star is found to be practically the same for both plates. This result 
indicates that, so far as these observations go, the probable error arising from the 
process of photographic reproduction is practically equal to the probable error in 
the position of a star image due to coarseness or irregularities in the film. 
7. A Comparison of the Long-period Rainfall Records at Cape Town and 
Greenwich. By HugH Rosert Miu, D.Sc. 
A cursory examination of the records of annual rainfall kept from 1841 to 
1904 at the Royal Observatory, Cape Town, and the Royal Observatory, Greenwich, 
shows a general similarity in the recurrence of series of wet. and dry years, not 
corresponding to definite cycles, but occurring in irregular spells of variable length, 
separated by unequal intervals. The total annual rainfall at the two stations does 
not differ greatly, but the range is greater for the Cape Observatory than for 
Greenwich, as shown in the following table :— 
—_— Cape | Greenwich | Cape | Greenwich 
in. | in, % % 
A. Mean : ; nite Vis = . | 25°94 | 24:14 100 100 
Be Weltesttyear~. te se ep ee | as 35°64.) bs) eee 
C. Driest ,, : - - : oul eekecOT 16:38) | 66 68 
D. Wettest three consecutive years ./| 34:42 30-01 iB By) 124 
E. Driest + * 43 ’ 18:94 20°71 73 86 
Difference B-C. ‘ , . | 23°96 19:16 | 92 79 
ss D-E. ‘ : : - | 15°48 9°30 60 38 
On a curve smoothed by taking the mean of three consecutive years, and 
plotting it on the place of the central year, it is seen that the wet spell 1858- 
1862 and the dry spell 1894-1901 were common to both, but that in almost all 
other cases a wet spell in the Cape peninsula corresponded to a dry spell in 
London, and vice versa. There was no apparent recurrence at regular intervals, 
and the remarkable flatness of the Greenwich curve during recent years had no 
analogy in the Cape curve. 
8. The Effect of the Sun-spot Period on the Daily Variation of the Magnetic 
Elements at the Cape of Good Hope. By G. H. Fincuam, B.A. 
Sabine’s results at the Cape of Good Hope for each of the years 1842-1847 for 
declination, horizontal intensity, and vertical intensity have been used to determine 
the daily variations in absolute units. 
