Page:Collected Physical Papers.djvu/85

is used, and the image is observed through a direct vision spectroscope, interference bands are observed in the spectrum, which flit across it with the slightest rotation of the cylinder.

I give below the results obtained from determinations made on different days (temp. = 26°).

(1) The larger apparatus was used for this experiment (diameter of the cylinder = 25 cm.), and the angle repeated fifteen times.

(2) Experiment with the smaller apparatus (diameter of the cylinder = 14·4 cm.). Reading was only taken of two successive positions for total reflection with the air-film 'direct' and 'reverse.'

(3) The same as above but with a different air-film.

(1) Angle repeated fifteen times.

Direct.⁠	⁠Reverse.
1459°	⁠1460°

Mean value for $\scriptstyle {i={\frac {1459\cdot 5}{2\times 15}}=48^{\circ }\ 39^{\prime }}$

μ = 1·3321.

From (2) and (3), mean value of i obtained = 48° 40′.

μ = 1·3317.

The values of μ from the best determinations reduced to the temp. 26° are given below:—

Gladstone and Dale •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •

1·3315

Terquem and Trannin •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •          •

1·3329

The result obtained by me is thus seen to be practically the same as the above.

Having obtained the absolute value of the index of water for the D line and at the temperature of 26°, the

G