222 
NATURE 
[Fan. 23, 1873 
The Twinkling of the Stars 
Every one who observes the stars at all must have noticed 
that they twinkle much more on some nights than on others, 
and this irrespective of any sensible difference in the clearness 
of sky or air. On rare occasions the twinkling becomes a really 
striking phenomenon, and at such times it is interesting to note 
the series of changes which together make up a ‘‘ twinkle.” 
For this purpose it is convenient to select two stars of suitable 
size and distance apart, and to look steadfastly at one, while 
the attention is directed to the other. The star which is not 
looked at will become alternately visible and invisible, and the 
manner in which these changes succeed each other will be found 
rather remarkable. 
On the evening of the rst of the present month, observing an 
unusual degree of twinkling, I made the above experiment on 
the stars e (Epsilon) and ¢ (Zeta) of Ursa Major. Looking 
steadily at either one of these, I noticed that the other, which 
was normally quite apparent, became every now and then totally 
invisible, and that not for an instant, but for a period of some 
duration, On one occasion I actually counted 30 in the interval 
of disappearance, and this I found afterwards to correspond to 
five seconds. More frequently, the star would be invisible for 
one or two seconds, then suddenly flash into full brilliancy, and 
after a variable interval vanish as suddenly again. 
From this it would appear that a ‘‘ twinkle,” at least when 
strongly marked, may be resolved into a sudden accession of 
brightness following a more or less prolonged period of com- 
parative obscuration. 
Stars may easily be found which will show the phenomenon I 
have described, even more strikingly than the two above named. 
I once tried two of the bright stars in Orion, and in this case 
the apparent sudden and absolute extinction, from time to time, 
of a conspicuous object, produced an effect almost startling. 
Clifton, Dec. 12 GrEoRGE F, BURDER 
Logarithmic Tables 
THE general procedure in determining numerical values in a 
scientific investigation is as follows. From a few observations 
we first compute the approximate values of certain constants, 
using for this purpose a theory which is purely a mathematical 
fiction ; and then, secondly, by comparison with extended series 
of observations we form equations of conditions, and determine 
the small corrections required by the approximate values of the 
constants. In the first part of this work logarithms of seven or 
more decimal places are necessary, but in the second part, which 
is generally by far the most laborious, logarithms of four and five 
decimals can be extensively used. Hence it is important that 
we should have well-arranged and convenient tables of such 
logarithms. An objection to nearly all the small tables that I 
have seen is that they are encumbered with tables that are not 
necessary to, or which do not properly accompany a table of 
logarithms, such as anti-logarithms, tables of meridinal parts, 
&c., and the result is that the logarithmic tables are made in- 
convenient for use. 
In the logarithmic tables recently edited by Prof. J. M. Peirce, 
(Ginn Brothers, Boston, 1871), the arrangement of the logarithms 
of numbers and of the Gaussian logarithms leaves nothing to be 
desired, and the method of printing the agreement in larger type 
is a good one. In his table of the trigonometric functions Prof. 
Peirce has also introduced a good idea in giving the double ar- 
gument, avz and ¢ime. This arrangement of the trigonometric 
function is however different from the one generally given, and 
hence for a computer accustomed to the common table is not 
convenient. I think that a table of four figure logarithms, in 
which the logarithms of numbers and the Gaussian logarithms 
should be printed after the arrangement and with the excellent 
type and paper adopted by Prof. Peirce, and with the trigono- 
metric functions arranged in the common order with the double 
argument a7c and ¢imz, and which should contain nothing else, 
is a desideratum. 
For tables of five decimal places I would follow the same order 
of arrangement, but would print the argument to the trigono- 
metric function in arc only, and would add a small table of 
squares for use in least square work. AsaPH HALL 
Washington, Noy. 9 : 
** Will-o’-the-Wisps ” 
Pror. GEIKIE, in his introductory lecture of the Murchison 
Chair of Geology at Edinburgh, which appeared in NATURE, vol. 
vii. p. 184, mentions that he never had the good fortune to en- 
counter one of these legendary sprites. It may not be unin- 
teresting to some of your readers to know that they are still 
extant. On October 5 last I was walking to the ‘‘ Lizard” with 
a friend, and near Ruan Major we saw a light travelling fast 
over the country, which my friend took to be the light of a dog- 
cart. As there was no road in the neighbourhood we watcha 
and soon saw two others rising from the same place and bound- 
ing over the country till they seemed to be about thirty feet from 
the ground in a swampy field opposite us, when they disap- 
peared. Another rose from the other side of the field, and after ~ 
reaching the middle of the field, it also disappeared. In about 
ten minutes we saw five or six, but none afterwards. 
I have asked several farmers of the district and many of my 
friends if they had ever seen any, but have only met with one 
farmer who said that when a boy he used to see them on Goon- 
hilly Downs adjoining. The geological formation of this district 
is serpentine. Howarp Fox 
Falmouth, Jan 15 
Spectroscopic Observations 
IN corroboration of Capt. Herschel’s statements regarding the 
mistaken idea of high dispersive power being essential to success 
in observations of solar prominences, I beg to give a few results 
obtained by a direct-vision spectroscope of dispersive power in- 
sufficient to separate D. 
An object glass of 2” diameter and 2’ 54” focal length (solar) 
was attached to this spectroscope in Jan last ; and on the 
first observation—using coloured glass that absorbed rays from 
B to a point rather less refrangible than F—the latter line was” 
found bright at four points on the sun’s periphery, the slit being- 
placed radial as well as tangential to the limb. 
Since then I have frequently observed prominences with and 
without the coloured glass, and on one occasion obtained G@ 
bright. In this case the prominence, which occurred on the day 
preceding the binocular eclipse of June last, was a small one, but 
C, the line near D, and F, were all intensely vivid. 
By the same spectroscope can be observed the brilliant lines 
of y Argus, as also the principal lines of a large number of stars, 
without using a cylindrical lens. 
At the red end of the spectrum I have obtained a broad belt 
of atmospheric absorption lines still less refrangible than the 
solar line that lies beyond the double atmospheric band on the 
red side of A. f ‘ 
I do not quite agree with Captain Herschel in attributing 
nothing to an Indian atmosphere, for the air here is doubtless 
more homogeneous than in the variable climes of Europe, but 
his protest against the prevalent notion of instruments of small 
dispersion being worthless for solar observations cannot be too 
widely circulated. 
Many valuable data have probably been lost to science by ob- 
servers being unaware of the power of the instruments at their 
disposal to work out the problems of nature. 
Mangalore, Nov. 26 E. W. PRINGLE 
GEORGE CATLIN 
R. GEORGE CATLIN, whose death we referred 
to last week, died in Jersey City on the 23rd of 
December last, after a lingering and painful illness. Mr. 
Catlin was born at Wilksbarre, Pennsylvania, on the 26th 
of July, 1796. 
Mr. Catlin began the series of Indian paintings which 
has made his name so well known everywhere, when 
accompanying Governor Clark, of St. Louis, in the years 
1830 and 1831, while he was engaged in making treaties 
with several Indian tribes. In 1832 he ascended the 
Missouri to Fort Union, and afterward returned in a 
canoe with two companions, a distance of 2,000 miles, 
visiting and painting all the tribes, so numerous at that 
time on the whole length of the river. Between this and 
1847 he made several extended journeys among various 
North American tribes, often sailing hundreds of miles in 
a bark canoe, ‘ 
By this means he accumulated a large number of 
paintings representing the portraits of the principal 
men of the tribes, and pictures of savage life, which 
were exhibited by him in various parts of the United 
ails 
