Construction of Wollafitotis Goniometer 217 



ed for, as the readings have always been taken on opposite sides 

 of the limb, so as to eliminate the error in centering, while no 

 errors need now-a-days be feared in a graduation to minutes. 



This apparatus is, as many trials have convinced me, quite 

 sufficient to give the inclination of two reflecting surfaces to the 

 nearest minute ; nor does it seem to me that, in order to ren- 

 der it still more precise, any thing else is required than the 

 addition of the telescope, and the use of more delicate gradua- 

 tions. Although a very convenient appendage to the work- 

 room, it is by no means adapted for the general purposes of the 

 crystallographer, who requires a portable and manageable tool. 



The manipulation of the common reflecting goniometer re- 

 sembles that above described in every thing but the steadiness 

 and certainty of the operations ; one source of inaccuracy so 

 exceeds all the rest, that it may serve at once to characterise the 

 instrument ; I mean the instability of the frame which carries 

 the reader : this results from the goniometer being detached 

 from the two parallel lines or objects of comparison, and from 

 its extreme lightness. Any change in position which occurs 

 during the observations induces an error in the results ; and 

 unless particular precautions, such as fastening the instrument 

 to the table, be taken, the measurements can never be entirely 

 depended on. 



It is by no means difficult to obviate this inconvenience, and 

 to give at the same time a compact and completely portable 

 form to the common goniometer. 



For this purpose, I fix upon the sole of the instrument a 

 small plate of brass, by means of two screws, one of them work- 

 ing in a round, and the other in an 

 elongated opening ; so that the brass 

 may have a limited motion round the 

 first, and may be clamped in any re- 

 quired position by the second screw. 

 To this I attach a small stage with its 

 upper face incled about 45, the particular inclination being of no 

 moment, and on that face I lay a piece of good thin plate glass. 



Instead of bringing the reflected light of an object to coincide 

 with the direct light of some other object, I cause it to coincide 



