DECEMBER 1, 1899. ] 
clamped to the objective end of the tele- 
scope bears a series of five slats, each 25 mm. 
wide, infront of the objective. These slats 
rotate about their longitudinal axes and are 
s0 connected as to turn together by means 
of a wire cord passing over a pulley near 
the eye end of the telescope. The entire at- 
tachment at the objective end is made of 
aluminium, and the total weight added to the 
telescope is 27 oz. (0.76 kg.). As viewed 
with the slats partly turned, a bright star 
appears as a central stellar image with a 
line of spectral images extending on either 
side at right angles to the direction of the 
slats. Comparisons indicate that this novel 
appearance has no influence upon the 
probable error of a single chronographic 
signal. 
E. E. Barnarp: Triangulation of Star 
Clusters. 
From the advantages given by the great 
scale of the 40-inch, it was decided to meas- 
ure the positions of a number of stars in 
some of the globular clusters, as a basis for 
the study of their motions in the future. 
For this purpose the clusters M 3, M5, M13, 
M 15 and M 92 were selected as representa- 
tive. 
In comparison with Scheiner’s photo- 
graphic measures of M 13 it was found that 
several of the stars had either diminished 
very greatly in brightness of late years, or 
that their light must come mostly from the 
blue end of the spectrum. The stars, 382 
and 393, of his list, are striking examples. 
Both these stars are given by him as 12.7 
magnitude. Visually, the first, which is 
his normal star, cannot be brighter than 
the 15th magnitude, while the second is at 
the limit of the great telescope and has 
been seen a few times only. Though it is 
possible that these stars may be variable, the 
observations do not seem to show it, except 
possibly in the case of No. 393. 
SCIENCE. 
789 
Variable Stars in Clusters. 
Some of the small variable stars discov- 
ered by Professor Bailey have been regu- 
larly observed during the measures of that 
cluster. Besides the small variables there 
are three rather bright stars in this cluster 
which are also variable. Two of these 
were discovered eight or ten years ago by 
Mr. C. D. Packer, of London. These three 
stars vary slowly, requiring upwards of a 
month for their variations, and one of them 
has a much longer period. The small stars, 
however, are very rapid in their changes, 
with nearly the same periods, averaging, as 
Professor Bailey has shown, about half a 
day. Variable No. 33, which rises from 
the 15th to about the 14th magnitude, and 
whose period is 12 hours 1 minute, has been 
specially observed at the request of Profes- 
sor Bailey. The observations show that 
this object lies dormant for a large portion 
of its period as a faint star of the 15th mag- 
nitude. It suddenly begins to brighten, 
and in about 15 minutes has doubled its 
light ; it then slowly declines to a position 
of rest. 
Several new variables were found in this 
cluster that are not marked on the Harvard 
photographs. 
The Fijth Satellite of Jupiter. 
The fifth Satellite of Jupiter was ob- 
served with the 40-inch, both in 1898 and 
1899, and good measures obtained of it. 
From the elongation distances obtained at 
these observations, a new value for the mo- 
tion of the line of apsides of the orbit was 
determined. This was found to amount to 
900° a year, or a complete revolution of the 
orbit in 4.9 months. With this value of 
the motion of the perijove, all the elonga- 
tion distances observed for the past seven 
years were accurately represented. A new 
determination was also made of the peri- 
odie time which is 11 h. 57 m. 22.6478. As 
this period was derived from nearly five 
