JiT.N-E 1, 1886.] 



♦ KNOAATLEDGE ♦ 



239 



render the spicier line in the e^-e-piece visible. Bright wires 

 would have been needed with any fainter light than this. 

 The rotating mirror was most ingenious, and Professor New- 

 comb expresses his obligation to Professor H. A. Rowland, 

 of Baltimore, for the valuable aid he rendered in devising it. 

 It consisted of a solid rectangular steel prism with carefully 

 groiind and polished sides. To the top and bottom of this 

 were ftistened pairs of circular plates, each pair holding a 

 set of twelve fans, upon which an air-blast impinged. The 

 top and bottom of the axes of the mirror terminated in 

 slightly conical pivots, the surfoce of the bottom one being 

 very nearly tiat and about '08 inch in diameter. This rested 

 on a diamond. By an elaborate system of airpipes the ob- 

 server could cause this mirror to rotate either to the right 

 or to the left, and by opening slightly the bottom apertures, 

 when the air-blast rotating the mirror was entering by the 

 top set, and vice versd, could regidate the rate of its motion 

 to a nicety. The number of complete rotations performed 

 in a second was measured by fixing a pinion of sixteen teeth 

 lound the upper axis of the minor. This geared into a 

 wheel with sixty-four teeth, as did that, in turn, by a ten- 

 tootli pinion into a wheel of seventy teeth, so that the latter 

 turned once for every twenty-eight times that the mirror 

 did. A knob in this last wheel at every revolution touched 

 a spring which formed part of an electric circuit passing 

 through a standard, so that circuit was bi'oken once in every 

 twenty-eight turns of the mirror. A cylinder chronograph 

 was caused to perform one revolution in ten seconds. Its pen 

 was connected with a closed circuit, passing through a break- 

 circuit chronometer and through the spring just spoi^en of. 

 Thus two series of indentations appeared in the line marked 

 by the pen on the chronogi-aph, the one at intervals of a 

 second and the other at the intervals in which the mirror 

 performed twenty-eight revolutions. A curious thing hap- 

 pened in connection with this recording apparatus. After 

 a few days' trial it was found that the wheel work was 

 entirely destroyed by the rapid rotation to which it was 

 subjected. New wheels wore out before it was practicable 

 to obtain a set of readings 1 In this extremity Professor 

 Newcomb returned the apparatus to its makers, Messrs. Chirk 

 and Sons, and thej' hit upon the ingenious device of employ- 

 ing raw hide for the fii'st wheel, with entire success. The 

 position of the receiving telescope was read off upon an arc, 

 whose centre lay in the prolongation of the vertical axis of 

 the rotating mirror, such reading being made by suitable 

 reading microscopes attached to the horizontiil frame carrying 

 the receiving telescope. The eye-end of the telescope itself 

 carried a micrometer. A little thought will suffice to show 

 that, as the transmitting telescope must send its i-ays ap- 

 proximately in the same direction as tiiat from which they 

 are received, if the two telescopes were directly over each other 

 the slit end of the transuiitting one would be light in the 

 way of the ob.server. This difficulty was overcome by the 

 simple expedient of bending the transmitting telescope at a 

 right angle, and putting a diagonal mirror at the bend to 

 reflect the light from the slit through the object glass. All 

 this will be rendered intelligible by reference to fig. 2. 



F is the sending telescope. The light of the sun is sent by 

 a heliostat through the slit S, passes down the tube and is 

 reflected by the diagonal mirror through the object-glass J. in 

 is the revolving mirror receiving this light and retlecting it 

 along Z to the distant fixed mirror (not shown). M,M,M,M 

 a stitf frame screwed into stone caps on the brick piers P,P' 

 with a horizontal divided arc on it at A. L the reeeivin" 

 telescope with its objective immediately below J, mounted 

 on adjustable Y's on the frame N which turns horizontally 

 round a vertical axis coincident with that of m. Its extreme 

 range of motion is about 8°. Tlie apparatus through 

 which the air-blasts pasts to turn the mirror m is omitted, as 



it is somewhat too complicated to be shown on the small 

 scale of our diawing. 



The measurements were performed by Professors New- 

 comb and Michclson, and bj' Ensign Holcombe, IT.S.N. The 

 stations were selected on the ojjposite sides of the Potomac 

 river ; the telescopes and rotating mirror being erected in 

 Fort Myer, on the Virginia side of the river, overlooking 

 the city of Washington. In the first instance the fixed 

 mirror was placed in the grounds of the Naval Observatory; 

 but, it being subsequently found that observations could be 

 made at a greater distance, this mirror was removed to a 

 point at the base of the Washington Monument. In the 

 first case the total distance between the mirrors was 1"585 



Ki(i 2. 



mile; in the second 2 .312 miles; the light, going and re- 

 turning, of course, travelling over double these distances 

 respectively. What we have called the fixed mirror was 

 really two mirrors of ir>| inches in diameter, placed side by 

 side. Tlieir radius of curvature was something like TSii 

 mile. 



As the I'esult of a most elaborate and exhaustive series of 

 observations with this beautiful apparatus, Professor New- 

 comb finally deduced as the velocity of light in vacuo 

 186,326±25 miles per second. It will interest the 

 astronomer to derive the solar parallax from this determina 

 tion, in combination -nith the now generally accepted 

 constant of aberration (that of Nyren) 20""-l:92, and 

 Colonel Clarke's computed equatorial radius of the earth, 

 3963-296 miles. It will be found to be 8"-794, corresponding 

 to a mean distance of the earth from the sun of 92,959,200 

 miles. 



At Professor Newcomb's request, Professor Michelsou 

 subsequently repeated his own determination of the velocity 

 of light by the method originally employed by him. As a 

 final result he obtained 186,322±37 miles per second as the 

 rate at which light travels in vacuo, an approximation sur- 

 prisingly close to that of Professor Newcomb, wheii we 



