32 
INATORE 
{ May 13, 1886 
fact, when reduced to a vacuum, 300,400 kilometres per 
second, while we may now regard it as well established 
that the true velocity is less than 300,000. 
The next determination of the velocity of light was 
that of Michelson,! whose result was 299,910 kilometres 
per second. His investigation being a part of the first 
volume of the present series need not be here discussed, 
but it is worth while to remark that his method seems 
far superior in reliability to any before applied. 
An attempt has been made by Messrs. James Young 
and George Forbes to improve Fizeau’s method, by 
diminishing the uncertainty arising from the gradual 
extinction of the visible image.’ By the method of these 
experimenters the result depends, not upon the moment 
when the image disappears, but when two images, side 
by side, are equal in brightness. This is effected by 
employing two reflectors, at unequal distances, but nearly 
in the same line from the telescope, to return the ray. 
Each reflector then forms its own image in the field of 
view of the sending telescope. With a regularly increas- 
ing velocity of the toothed wheel, each image goes inde- 
pendently through the same periodic series of changes as 
when only one mirror is used ; but owing to the unequal 
distance the period is not the same. If the speed of the 
mirror be carried to such a point that the difference of 
phase in the two images is half a period, then one image 
will be increasing while the other is diminishing, and the 
stage at which the two images are equal would appear to 
admit of fairly accurate determination. 
The distant reflectors were separated from the observing 
telescope by the Firth of Clyde. The distances were 
respectively 16,835 feet, and 18,212 feet. A study of the 
printed descriptions of their experiments gives the im- 
pression that the performance of the subsidiary parts of 
the apparatus was not suchas to do justice to the method. 
The resulting velocity of light was 301,382 kilometres 
per second, and the difference between the extreme 
results of twelve separate determinations was 4000 
kilometres. 
‘The most important result of the work of these gentle- 
men, could it be accepted, would be the establishment of 
a difference between the velocities of differently-coloured 
rays. We may regard it as quite certain, from the 
absence of any change in the colour of the variable star, 
8 Persei, while it is increasing and diminishing, that the 
difference between the times required by red and by blue 
rays to reach us from that star cannot exceed a moderate | 
fraction of one hour. It is quite improbable that its par- 
allax is more than o’1, and therefore probable that its 
distance is 2,000,000 or more astronomical units. The 
possible difference between the velocities in question can, 
therefore, only be a small fraction of the hundred-chou- 
sandth part of either of them. No apparatus yet devised 
would suffice for the measurement of a difference so 
minute, and we are justified in concluding that the phe- 
nomena observed by Messrs. Young and Forbes arose 
from some other cause than a difference between the 
velocities of red and blue rays. 
The present determination had its origin as far back as 
1867. In his “Investigation of the Distance of the Sun,” 
published in that year, the author introduced some 
remarks upon Foucault’s method, and pointed out the 
importance to the determination of the solar parallax of 
repeating the determination of Foucault on a much larger 
scale, with a fixed reflector placed at a distance of three 
or four kilo netres.* 
From that time forward the subject excited the atten- 
tion of American physicists, several of whom formed 
plans, more or less definite, for executing the experiments. 
As, up to the year 1878, no important steps in this direc- 
7 “ Astronomical Papers of the American Ephemeris,”’ vol. i. partiii. Owing 
to anerror in applying one of the corrections the result was given as 299.042 
kilo netres. 
? Philosophical Transactions {or 1882, p. 231. 
3 Washington Jbservations, 1865, Appendix ii. 
tion had been taken, the author, in April of that year, 
brought the subject before the National Academy of 
Sciences, with the view of eliciting from that body an 
expression of opinion upon the propriety of asking the 
Government to bear the expenses of the work. The 
subject was referred to a Select Committee, who,in January, 
1879, made a favourable report on the subject, which was 
communicated to the Secretary of the Navy. On the 
recommendation of the Secretary, Hon. R. W. Thompson, 
Congress, in March following, made an appropriation of 
$5000 for the purpose, and the author was charged by the 
Department with the duty of carrying out the experi- 
ments. 
In the meantime it became known that Mr. Michelson 
had made preparations for rep2ating Foucault’s deter- 
mination at his own expense, with the desirable improve- 
ment of placing the fixed reflector at a considerable dis- 
tance. But before the reliability of Mr. Michelson’s work 
had been established, the preparations for the present 
determination had been so far advanced that it was not 
deemed advisable to make any change in them on account 
of what Mr. Michelson had done. The ability shown by 
the latter was, however, such that, at the request of the 
writer, he was detailed to assist him in carrying out his 
own experiments, and acted in this capacity until Sep- 
tember 1880, when he accepted the Professorship of 
Physics in the Case Institute, Cleveland, Ohio. After the 
departure of Mr. Michelson his place was taken by 
Ensign J. H. L. Holcombe, U.S.N., who assisted in 
every part of the work to the entire satisfaction of the 
projector until its close. 
PANCLASTITE 
R. SPRENGEL has sent us a reprint of a note sent 
by him to the Chemzca/l News on this subject. After 
showing that these new explosives, so named by Mr. Turpin, 
are not original, he continues:— _ 
“The ‘beau idéal’ of a detonating explosive is a mixture 
of 8 parts (88°9 per cent.) of liquid oxygen and I part (11°1 
per cent.) of liquid hydrogen. 
“In my paper of 1873 I say, p. 799 :—‘ On referring to 
the foregoing table the reader will be reminded that per- 
oxide of hydrogen is the highest oxygen compound known, 
while nitric anhydride is the compound which contains 
the largest amount of oxygen available for combustion 
(74 per cent.). But as this compound, as well as the next 
two, nitric peroxide (69°5 per cent. oxygen) and tetranitro- 
methane (65°3 per cent. oxygen) are ; at present: on 
account of their nature and their difficult preparation, 
mere chemical curiosities, my attention naturally turned 
to the fourth, to z¢r?c acd (63°5 per cent. oxygen), which 
is a cheap and common article of commerce.’ : 
“ Now, when Mr. Turpin’s attention turned to the second 
oxidiser on my list—to nitric peroxide—he found that this 
substance does wo¢ corvode metals, such as iron, copper, 
and tin under 356° F. (180° C.) ; and further, that com- 
bustible liquids, such as petroleum, carbon bisulphide, and 
nitro-benzene are readily soluble in nitric peroxide we¢h- 
out rise of temperature. These are valuable properties, 
Jirst noticed by Mr. Turpin. 
“What was formerly a chemical curiosity is now an 
article of commerce. Nitric peroxide may be bought 
to-day at eighteenpence the pound, and I see ways and 
means of producing it a great deal more cheaply. Nitric 
peroxide is a yellowish liquid, heavier than water (sp. gr. 
= 1°451), and boils at 71° F. (22° C.), but may be kept like 
ether or similar volatile liquids. In France itis sent about 
in tinned-iron cans. 
“ Taking as a typical example a benzene-mixture— 
CyHy = 184 = 7° | _ 626 CO, 
568 oJ Wi2zoH sO 
74(N Oz) _— = 34°8N 
