NATURE 



\yuly 19, 1877 



electro- magnet, the tracer makes a short mark at right 

 angles to line, and a zig-zag line caused by the vibra- 

 tion of the tracer, back to its original position ; the first 

 two lines, therefore, show seconds and tenths of seconds, 

 the third, the instants of completion of forty or 400 

 levolu'ions, according to the desire of the operator, of the 

 toothed wheel ; a comparison, therefore, at once gives the 

 number of revoluiions per second, while on the fourth 

 line are marked the instants of disappearance or reap- 

 pearance of the light, and the velocities at those instants 

 are then at once known. To make an experiinent the 

 aperture in the diaphragm is illuminated by a lime-light 

 or sometimes with sunlight by means of a heliostat. The 

 necessary adjustments in the direction of the rays of light 

 to the disiant station are then made by bringing the 

 distant collimator into the centre of the field of the ob- 

 serving apparatus, and the point of light — the luminous 

 tcho — is made to accurately coincide with its original at 



-P:ao of M. Fiz 



Apparatu 



the circumference of the toothed wheel. Particulars of 

 the experiment, as to number of teeth of wheel, diiection 

 of rotation, &c., are entered on the paper on the cylinder, 

 and the latter is then set in motion ; the observer then 

 sets the toothed wheel going and watches the luminous 

 echo, and on its disappearing touches a key which sends 

 an electric current to the electro-magnet controlling the 

 fourth tracer, which therefore registers the instant the 

 velocity is sufficient to cause a disappearance. As the 

 velocity of the wheel increases, the luminous echo again ap- 

 pears and the key is pressed ; a further increase in velocity 

 causes another disappearance and so on to the higher 

 orders, each of which is registered. The velocities at the 

 different instants are read off by a micrometer to ^ Jj of a 

 second. 



It is obvious that the state of the air must have a great 

 effect on the definition of the luminous echo, and that 

 although the observation appears extremely simple still 



there may be large errors due to irregular refraction of 

 the air, causing a motion of the point of light, and a large 

 amount of patience must be required. Two carelul 

 surveys showed the distance between the two stations to 

 be 2290977 metres, and the mean velocity obtained from 

 a large number of observations after the various correc- 

 tions were made was 300,400 metres per second of mean 

 time. 



The Memoir of M. Cornu contains a large amount of 

 theoretical matter and formula: of corrections which of 

 course we cannot reproduce here. 



We may, however, refer to the principal causes of error. 

 The first is a personal error depending on the sensibility 

 of the eye of the observer in determining the disappear- 

 ance and reappearance of the light at the toothed wheel, 

 and also depending on the intensity of the luminous 

 source ; secondly, accidental inequality in the size of the 

 teeth of the wheel ; thirdly, irregularity of motion ; 

 fourthly, excentricity of wheels ; fifthly, optical errors due 

 to imperfections in the adjustment of the lenses and re- 

 flector. The first of these is small and can theoretically 

 be reduced indefinitely by increasing the velocity of the 



Fig. 2. -Details of Tooilieu Wheel, 



toothed wheel and thereby observing the higher orders of 



extinction and reappearance of the light, but M. Cornu 



1 desires to be rigorously exact, and therefore the effect of 



this and the other errors is carefully calculated. 

 I Considerable care was exercised in the choice of sta- 

 tions, and those adopted were fixed upon chiefly on ac- 

 count of the distance between them being more easily 

 ascertained from previous triangidation. This distance 

 i was determined by Cassini and La Caille in 1740, the 

 : result being 229io'ig6 metres. From the observations of 

 j Delambre the same distance was computed to be 22909'34 

 metres ; and 22910 metres, which is nearly the mean 

 ; adopted by M. Cornu. 



The corrected results of the experiments gave a velocity 

 ' of 300,350 kilometres a second, but this was obtained in 

 I air, and therefore 82 kilometres must be added to this 

 1 result to give the velocity in vacuo ; and as the result of 

 I his experiments M. Cornu adopts a velocity in vacuo of 

 ' 300,400 kilometres = 186,638 miles per second of mean 

 I time, with a probable error of i Asi or 300 kilometres. 

 j If from this value we deduce the solar parallax, we find 

 ' the latter to be 8"88r, assuming the time required for light 

 ' to travel from the sun to us to be Cm. i3'2 sec, as obtained 



