176 PEOCEEDINQS OF THE [AugUSt, 1887. 



tant horizon glasses answer very well,) should be cemented together, the re- 

 flecting surface of each being perpendicular to that of each of the others. It 

 may be shown that the light from the sun, reflected successively from two of 

 these plates, and passing through the third, will be moving parallel with, but 

 opposite to the motion of the light reflected from this third surface, and so will 

 the problem above be solved. I have constructed such an instrument and used 

 it successfully for short distances, but have not tried it for long ones, as there 

 will be uncertainty about the accuracy of the angles, unless caxefully tested or 

 measured. This instrument is more compact than any of the others, and might 

 consist of a single piece of plate glass, as Mr. MiUer has shown, by having two 

 additional planes ground at one of its corners, and carefully polished, so as to 

 have three planes accurately placed at right angles to each other. 



The last two instruments are, as has been shown, eminently portable, but 

 nothing forbids their being mounted on stands when convenient to do so, as 

 in making use of the Morse alphabet. 



Of all these instruments, Galton's is the easiest to construct, and requires the 

 fewest adjustments, as is evident from the instruments I have now exhibited. 



SEPTEMBEK 29th, 1887. 

 The President in the Cliair. 



BooJcs Received. 



Naturforschenden Gesellschaft, Gorlitz : Abhandlimgen, Band 

 XIX. 



Royal Society, Dublin : Scientific Transactions, Series II, Vol. 

 in, parts 11-13, and Vol. V, parts 3-6. 



Museo Nacional de Mexico: Anales, Tomo 3, Entr. 11. 



Smithsonian Institution : Report, 1885, Part I. 



Museum of Comparative Zoology, Cambridge : Bulletin, Vol. 

 XIII, No. 4. 



Royal Society, London : Proceedings, Nos. 255-6. 



