292 



"4. It certainly produces some optical indistinctness, which 

 at so great a distance I did not expect. At first a train passed 

 at noon ; and we were surprised to find that the suns limbs, 

 though perfectly well defined at the first wires, would some- 

 times suddenly, as it were, shiver and become confused so as 

 to preclude all observing ; a little attention to the steam- whis- 

 tle explained its cause, and had not the train been suppressed, 

 we must have neglected solar observations for half the year. 

 The effect was similar in both instruments, but I think greater 

 in the circle. With respect to stars, it was chiefly studied on 

 the pole-star, whose slow motion gave ample leisure to exa- 

 mine the appearances. In general, the star became a luminous 

 blot of an elliptic form ; but when the air was perfectly still, 

 so that the definition was perfect, the phenomena were much 

 more striking. In such cases the star is a mere point sur- 

 rounded by a system of coloured rings ; these, about five mi- 

 nutes before the train came in, elongated themselves with rapid 

 oscillations nearly in the meridian. In another minute the cen- 

 tral point began to dilate, and as it met the lesser axis of the 

 distorted rings they coalesced, till all became a bright blur, as 

 large as the inner ring, in two minutes more. As the speed 

 was diminished, traces of the rings re-appeared, and when the 

 train stopped, all was as at first. As respects stellar astronomy, 

 this is not of much importance, for it will cause the loss of only 

 two or at most three observations for each train, and only a 

 rare chance could make it interfere with an occultation ; with 

 the sun it is otherwise, but I trust that there are no railway 

 directors who would not in such a case alter their arrange- 

 ments, so as to leave the 20 m before and after noon free. 



" I will conclude by suggesting whether, henceforward, it 

 will not be prudent, in selecting sites for observatories, to choose 

 with special reference to the ground's capability of conducting 

 tremors, and ascertain it by trial. From what I have stated, 

 it seems probable that a deep and homogeneous substratum is 

 in this respect the very worst ; while the beautiful experiments 



