156 ON THE HORIZONTAL MOON. 



Properties of that, on account of the fhortnefs of this bed, it could not pof- 

 pendulums, Sec. fibl_y expand as much in length as the 39 inches of wire con* 

 tained in a common pendulum-rod ; and that, as the greateft 

 expanfion of the wire is certainly in the direction the moil in- 

 jurious to the effect of the pendulum, which is not the cafe 

 with that of the cycloidal bed, it is probable that the rolling 

 pendulum would be found fuperior to that in common ufe. 



With regard to the oblervations on pyrometers, I own I 

 never faw that of Deluc ; but as your defcription of it is not 

 fufficiently minute to enable me to perceive how the ftandard 

 in it for meafuring the different bar?, could efcape expanfion 

 in an increafe of temperature of the atmofphere, I ftill can 

 conceive no method by which the ftandard for meafurement in 

 any pyrometer, could be prevented from altering its length 

 from the above caufe, but by keeping it artificially at a fixed 

 temperature, by fome means diftinct from the apparatus, limi- 

 lar to thofe of Mr. Ramfden, which you have mentioned. 

 What I wrote on this fubjeft was intended as a caution in 

 future experiments, and if it has produced the more extended 

 publication of means to avoid the error I pointed out, or (hall 

 hereafter do fo, my view will be fully anfwered. 



Your very humble fervant, 



J. WHITLEY BOSWELL. 



HI. 



, Some Remarks upon the Experiments bytihich Mr. Ez. Walker 

 has endeavoured to explain the apparent Enlargement of the 

 Moon near the Horizon : with a Statement of fome Facts upon 

 which that Phenomenon feems chiefly to depend. In a Letter 

 from C. H. 



To Mr. NICHOLSON. 

 SIR, 



Mr. Walker's A SHOULD, with your permiffion, be glad to point out to 

 experiments ad- your correfpondenf, Mr. Ezekiel Walker, a theorem in op- 

 tics, to which, it appears to me, that he has not paid fufficient 

 attention in the courfe of his ftudies. 



Rays 



