380 SURVEY OF THE COAST 



These light parts are very liable to spring back, on being 

 exposed to a considerable variation of temperature. By this 

 construction, Mr. Reichenbach has applied verniers in his 

 circle of the Paris Observatory, of which the diameter is one 

 metre, and which is the largest size he finds advisable to 

 give to any instru.nent. 



The English artists, on the contrary, use compound mi- 

 croscopes for the readings, where the diameters are above 

 eighteen inches, atd find them to be very convenient. These 

 microscopes, and probably the readings with them, are more 

 accurate ; but they require more labour and very accurate 

 work. 



I have already observed, that for a division by points, the 

 micrometer must have one single wire in the direction of 

 the radius ; and for a division by lines, two wires, intersect- 

 ing each other and inclined equally to the radius, making 

 angles of about fifteen degrees on each side ot the division 

 stroke, the equality of which is very easily judged with great 

 accuracy. 



The divided circle, as well as the radii, arms, or plates 

 bearing'verniers or microscopes, ought never to be attached 

 to other parts of the instrument, except the central piece, 

 in order to be in all cases equally free, and left to their own 

 spring and shape. In like manner the centre piece is to be 

 connected with the stand part alone. 



The clamping and tangent screw arrangement ought al- 

 ways to be strong, and never to be fixed to an arm bearing 

 a vernier or a microscope. It is very apt to affect it differ- 

 ently from the other verniers, as it has to act on all parts as 

 leader, and to overcome the resistance of the central friction. 

 On this account, it is commonly made somewl)at stronger 

 than the others. If tlie verniers are not attached to a full 

 plate, it will be best to give the instrument a separate strong 

 arm, expressly for the purpose, and to make all arms of ver- 

 niers or microscopes exactly equal. 



The quality and power of the telescopes on an instrument 

 must always exceed the degree of accuracy which is obtain- 



