Fune 24, 1886 | 
ceeding 230 revolutions per second,! gave a total change 
of direction of nearly 8°. And this largeness of the 
measured angle materially contributed to enhance the 
accuracy of the results. Highly effective, also, for the 
same end were the elaborate precautions for darkening 
the telescopic field of view, and thus rendering the image 
of the illuminated slit more distinct. As their upshot, 
daylight was reduced to about one-thousandth its normal 
intensity. What was left only just sufficed to show the 
spider-lines without artificial light. The necessity for 
such precautions may be estimated from our author’s 
statement that a concave mirror, of which the diameter 
should be one decimetre for each kilometre of distance, 
would receive only 1/60,000 part of the light reflected from 
the revolving mirror; while of that 60,oooth part only 
a small fraction could be practically turned to account, 
owing to the many sources of loss in reflection and trans- 
mission. Since, however, ¢wo fixed mirrors, each four 
decimetres across, placed at a distance of less than 
four kilometres, were employed in the operations at Fort 
Myer, the proportion of light there returned was rather 
more than double the above estimate. Prof. Newcomb 
appears to have been, on the whole, eminently successful 
in his optical arrangements. The imperfect definition 
which was the besetting difficulty of Michelson’s experi- 
ments gave him little trouble. 
The recent American determinations of the velocity of 
light, justly considered as of far superior precision to any 
others yet executed, give the following results :— 
Michelson, at Naval Academy, in 187 
Michelson, at Cleveland, 1882 ... ... 
Newcomb, at Washington, 1882, using 
only results supposed to be nearly 
free from constant errors... ...  ... 
Newcomb, including all determinations 
299,910 km. 
299,853 5 
299,860 ” 
299,810, 
To these are added for comparison :-— 
Foucault, at Paris, in 1862 ... 298,000 ,, 
Cornu, at Paris, in 1874 298,500 ,, 
Cornu, at Paris, in 18/8 Snr 300;4 00Kmmrs 
The same, discussed by Listing... ... 299,990 ,, 
Young and Forbes, 1880-81... ... 301,382), 
Prof. Newcomb’s finally-concluded result is that light 
travels 7 vacwo at the rate of 299,860 + 30 kilometres 
per second. And the probable error of thirty kilometres, 
small as it is, has been liberally estimated. A determina- 
tion so satisfactory of this important element goes far 
towards solving the problem of the sun’s distance. Com- 
bined with Nyrén’s constant of aberration, 20’°492, it 
gives, for the solar parallax, the value of 8794. The 
corresponding distance of 149°61 million kilometres, or 
2,965,020 miles, agrees quite closely with Dr. Gill’s result 
from the opposition of Mars in 1877, and exceeds by only 
165,020 miles the mean deduced by Mr. D. P. Todd from 
earlier determinations of light-velocity. No information 
as to the dimensions of the solar system which we are 
ever likely to get from a transit of Venus can approach 
in reliability the present conclusion. 
Prof. Newcomb is so far from believing that the 7e plus 
ultra of accuracy has been reached in his own remarkable 
experiments, that he appends to the detailed description 
of their method some valuable suggestions for its im- 
provement. He had hoped, indeed, he tells us, to reach 
a concluded value exact to between five and ten kilo- 
metres, which, after repeated verification, might be avail- 
able as a test of the invariability of standards of length. 
The further prosecution of the inquiry, however, he now 
leaves to any physicist who may be invited to the task by 
the promise of his advice and co-operation. 
Fundamentally, he holds that the system pursued at 
Fort Myer in 1880-82 is preferable to any other yet 
tried. No known expedient for ascertaining the rate of 
* Michelson’s revolving mirror executed 256 turns in a second. 
NATURE 
173 
transmission of light is comparable to that of its deflec- 
tion, after a measured journey, by a moving mirror. The 
apparatus by which this plan was realised admits, how- 
ever, in his opinion, of some amelioration in detail. The 
disadvantageous necessity, for instance, of appropriating 
a separate section of the reflecting surface to the outward- 
and homeward-bound rays could be removed by the sub- 
stitution of a pentagonal for a quadrangular prism, as 
shown in Fig. 3, where M is a section of the revolving 
mirror, J the object-glass of the sender, receiving light 
from the slit s, and throwing it in the direction P towards 
the distant reflector. On its return along the path P’, the 
ray is reflected from an adjoining face of the revolving 
mirror into the receiving telescope, K. 
The closing words of the paper under review attest the 
unappeased aspiration towards accuracy characteristic of 
the successful investigator. 
“A still further perfection of the method,” its author 
writes, “which would lead to a result of which the pre- 
cision would be limited only by our means of linear 
measurement is, I conceive, within the power of art. It 
consists in placing the fixed mirror at so great a distance 
that the pentagonal revolving mirror would move through 
an arc of nearly 36° while the ray is going and returning. 
If a speed of 500 turns per second could be attained, 
the required distance would be thirty kilometres. Then, 
in opposite directions of rotation, the return ray would 
be reflected at phases of the mirror differing by the angle 
between two consecutive faces. The result would be that 
the receiving telescope would need to have but a small 
motion, and all the observer would haveto measure would be 
the small angle by which the difference of positions of the 
mirror when the flash was received in opposite directions 
of rotation, differed from 72°. In the Rocky Mountains 
or the Sierra Nevada no difficulty would be found in 
finding stations at which a return ray could be received 
from a distance of thirty, forty, or even fifty kilometres, 
with little more dispersion and loss than ata distance of 
four kilometres through the air of less favoured regions. 
It is true that the surface of the distant reflector would 
have to be increased in proportion to the distance, but it 
would not be necessary to make a single reflector of great 
size. A row of ten reflectors, each six or eight decimetres 
in diameter, might be sufficient to insure the visibility of 
the return ray.” A. M. CLERKE 
NOTES 
At a meeting of the Royal Society of Edinburgh on June 7, 
medals were presented as follows:—To Mr, John Aitken 
(Darroch), the Keith Prize for 1883-85, for his paper on the 
formation of small clear spaces in dusty air, and for previous 
papers on atmospheric phenomena ; to Edward Sang, LL.D., 
the Makdougall-Brisbane Prize for 1882-84, for his communica- 
tion on the need for decimal subdivisions in astronomy and 
navigation and on tables requisite therefor, and generally for 
his recalculation of logarithms both of numbers and of trigono~ 
metrical ratios ; to Mr. B. N. Peach the Neill Prize for 1883- 
86, for his contributions to the geology and paleontology o 
Scotland. 
THE organising committee of Section A has arranged that a 
special discussion shall be held, at the Birmingham meeting of 
the British Association, jointly with Section D, on the physical 
and physiological theories of colour-vision. The discussion will 
be opened by Lord Rayleigh, and Dr. Michael Foster will also 
take part in it. Persons who wish to contribute papers bearing 
on the subject of discussion are requested to send their names to 
the Recorders of Sections A or D, at 22, Albemarle Street, W., 
not later than August I. 
THE death is announced, in his seventieth year, of Mr. 
Llewellyn Jowett, the well-known archeologist. 
