168 Suggestions for the Observation 



Journal, vol. xx., p. 83, and in the Philosophical Magazine, 

 vol. iv., p. 289 ; and instruments on this construction are sold 

 by M. Soleil in Paris, and Watkins and Hill in London. 



Savart's polariscope is described in Peclet's Traite de 

 Physique^ and is sold by the same artists. 



The accompanying map has been constructed principally 

 from computations furnished to the Committee by Lieut. W. 

 S. Stratfoi^i, R.N., Superintendent of the Nautical Almanac, 

 verified in some parts by duplicate computations made under 

 the direction of the Astronomer Royal. 



The elements employed for computation of the geocentric 

 places of the sun and moon are those of the Nautical Almanac. 

 The sun's semidiameter, as given in the Nautical Almanac, 

 is increased by tttVo part, the moon's parallax by 12^0 o part, 

 and the moon's semidiameter by g^^ part, in conformity with 

 the results of extensive investigations by the Astronomer 

 Royal. It is to be remarked that the semidiameter thus 

 found for the moon is that corresponding to an illuminated 

 moon seen on a dark sky : if the apparent semidiameter when 

 the dark moon is seen on the sun's bright disc be sensibly 

 smaller, the breadth of the shadow and the duration of total 

 darkness will be less than those given in the map. 



The numbers in the 1st, 2nd, 3d, 7th, and 8th columns are 

 computed for the points opposite to them in the central line 

 of shadow, but they will apply with sufficient accuracy for 

 other neighbouring points within the shadow. The numbers 

 in the 4th, 5th, and 6th columns are also computed for the 

 points opposite to them in the central line of shadow ; but 

 they require large corrections to make them applicable to 

 other points within the path of the shadow but not on its 

 central line. These corrections are given by the numbers 

 at the top and bottom of the map, corresponding to the 

 various lines drawn longitudinally through the shadow's 

 path. An example will best illustrate the mode of finding 

 these corrections. 



It is required to find the duration of total darkness and the 

 angles from the upper point of the sun's disc for disappear- 

 ance and reappearance, at Vladimir. 



Opposite Vladimir, the duration of total darkness on the 



