DeEcEMBER 10, 1896] 
NATURE 137 
o 
fied and extended. Without increasing its bulk, seventy-two 
new plates have been added ; the text has been rewritten, and 
the maps revised to the date of issue. —‘‘ A Text-book of Shades 
and Shadows and Perspective,” prepared for the use of students 
in technical schools, by Prof. John E. Hill, has reached a second 
edition. The book, which is published by Messrs. John Wiley 
and Sons, has been thoroughly revised and, in the main, re- 
written.—Messrs. Cassell and Co. have commenced the issue of 
a cheap edition of their ‘‘ Technical Educator.” The parts will 
appear weekly, and each will contain ninety-six pages of text and 
illustrations. 
THE additions to the Zoological Society’s Gardens during the 
past week include two White-thighed Colobus (Colobus veller- 
osus, & 9), a Campbell’s Monkey (Cercopzthecus campbelli, 6), 
a Green Monkey (Cercopithecus callitrichus,é) from West 
Africa, presented by Dr. S. H. Armitage; a Lesser White- 
nosed Monkey (Cercopithecus petaurista) from West Africa, a 
Black-headed Lemur (Lemz drunxneus) from Madagascar, pre- 
sented by Miss Baird; an Arabian Gazelle (Gazella arabica, & ) 
from Arabia, presented by Mr. R. G. Buchanan; two Rac- 
coons (Procyon Jotor) from North America, a Golden Eagle 
(Aguila chrysetus), a Buzzard (Buteo vulgaris), European, pre- 
sented by Lord Arthur Cecil ; two Laughing Kingfishers (Dace/o 
gigantea) from Australia, presented by Mr. F. Beaumont ; five 
Eyed Lizards (Lacerta oce//ata), European, presented by Mr. 
P. Goupille ; a Thick-necked Tree Boa (Zfecrates ceuchris) 
from Tropical America, deposited. 
OUR ASTRONOMICAL COLUMN. 
LEonID METEORS IN AMERICA.—The display of meteors on 
November 13 and 14 is said to have been great at some places 
in America. Published accounts include the statement that 
as many as twelve at a time were seen; anda press dispatch 
from Indianopolis states that meteors were visible there in 
broad daylight, in fact about the middle of the day. 
A New SPEcrRoscopic BINARY IN Puppis.—A new addi- 
tion has recently been made to that class of stars of which 
¢ Urse Majoris, 8 Aurigze, and y’ Scorpii are the present 
members. The star in question is Lacaille 3105, A.G.C.10534, 
and the Draper Memorial photographs show that it is a binary, 
its photometric magnitude being 4°50 (Harvard College Obs. 
Circular, No. 14). Prof. E. C. Pickering first noticed its 
binary character in February of last year, and this has been 
corroborated by several photographs secured by Prof. Bailey. 
A discussion ofall the photographs taken, gives the length of the 
period as 3d. 2h. 46m., the formula representing the times of 
inferior conjunction being given as J.D. 2412777.16 + 3°115E. 
At these times the lines in the spectrum are said to be single ; 
for the next thirty-seven hours they are double, the fainter com- 
ponent of each having a greater wave-length than the brighter 
component. The lines then again become single, and finally 
double for the remainder of the period, the fainter components 
having shorter wave-lengths, and being therefore towards the 
violet end of the spectrum. 
RELATIVE MOTION IN THE LINE OF SiGHT.—In a Har- 
vard College Observatory Circular (No. 13), Prof. E. C. 
Pickering describes a new method by which the relative motion 
of stars in the line of sight may be determined by means of 
the objective prism. As yet only preliminary trials of this 
method have been made, but it promises, as we are told, to be 
fairly satisfactory. Ina discussion with Mr. E. S. King, of the 
experiments made by comparing the corresponding lines in the 
spectra of different stars with their images taken on another 
plate without the prism, and with the film reversed, the follow- 
ing process was evolved. It may be briefly summed up as 
follows. Two stars which are close enough to be photographed 
ona single plate are first selected. Let one of these be supposed 
at rest, and the other moving, say, towards the observer with a 
velocity represented by a deviation of amount d in its spectrum. 
If a photograph be taken such that the end of the shorter wave- 
ength of the moving star be turned towards that of the star at 
NO. 1415, VOL. 55] 
rest, the distance between the images of a given line in the 
two spectra will be lessened by the amount ¢@. By turning 
now the prism through 180°, the spectra will also be rotated 
through the same arhount, and the end of greater wave-length of 
the spectrum of the moving star will now be turned towards 
that of the star at rest. If a photograph be taken in this position, 
the distance will in this case be increased by d. By superposing 
the two photographs thus obtained, making the images of the 
given line in the spectra of the star at rest coincide, the 
difference in the readings of the same line in the other spectra 
gives double the displacement of this line. 
In actual practice the photographs can be obtained one on 
each side of the meridian, the act of reversing the telescope 
turning the prisms, and with them the spectra, through 180°. 
For the examination of the plates, the film sides of the nega- 
tives must be placed together to ensure an accurate comparison, 
thereby necessitating that one of the photographs has to be 
obtained with the film side away from the objective prism. 
The question of change of focus can here then arise, of which, 
however, Prof. Pickering makes no mention. ‘ 
The advantages of this method, as stated by him, may be 
condensed as follows :—Directness of the determination of the 
motion. Double the deviation is measured. Accidental errors 
of measurements are much less than when each in turn is bisected 
by a spider line, as the ends of two similar lines are made to 
coincide. Since each line in the spectrum may be used, a large 
number of independent determinations may be obtained from 
one pair of plates. Further, a visual telescope may be employed, 
as it is only zecessavy that one line should be in focus. Cor- 
rections for the motion of the sun in space, or of the earth in 
its orbit, may be disregarded, as they affect both stars equally. 
FORMULA FOR COMPUTING WAVE-LENGTHS.—By the in- 
vestigations of several workers the positions of the lines in the 
spectra of numerous substances are being mathematically de- 
termined. It was Lecoq de Boisbaudran who first pointed out 
the regularity in position in which lines of some substances ap- 
peared. Stoney, and others after him, found out that some of 
the lines of hydrogen were harmonically situated in the 
spectrum. It was left, however, for Balmer to bring forward a 
formula by which the wave-lengths of the lines in the hydrogen 
spectrum could be computed with wonderful accuracy. The 
formula was as follows :— 
where 7 represented the numbers 3 to 15, and A the number 
3645°42 in Angstrém’s units. Cornu at the same time, and 
Deslandres a year later, worked on this question, the latter pro- 
ducing a formula from which the bands of several elements could 
be computed. 
In 1887 Kaiser and Runge set to work, and between them 
they advanced the new idea very considerably by describing a 
general formula which represented a great number of series of 
different elements, and which also included that of hydrogen, 
Balmer’s formula being found to be only a special case. The 
new formula became then 
T= A — Bu-? — Cx-4, 
the reciprocal of the wave-lengths (t,), instead of* the wave- 
length themselves being here employed. The first three members 
of ‘such a series sufficed to represent the series of lines with con- 
siderable approximation from values obtained from observation, 
with the exception, perhaps, of the longest and shortest waves. 
For the determination of these constants, which appear in this 
formula, three distinct measurements of the number of oscilla- 
tions are necessary. 
The drawback of this formula is that for the longest, or, at 
any rate, the two longest waves, it is not sufficiently accurate ; 
and, even if more terms were added, the determination of the 
values of the additional constants would be uncertain owing to 
the limited accuracy of the wave-length measurements. 4 
In determining the accuracy of this expression, Messrs. Kaiser 
and Runge were thus led to undertake the investigation of more 
accurate determinations of the spectra of the elements. | ; 
More recently, however, Herr Balmer has succeeded in finding 
an expression which represents the series with greater accuracy 
than formerly. This differs slightly from those given above, in 
that the third term is dispensed with, and an additional constant 
