Instrument for measuring the expansion of Solid Bodies. 329 



The thermometer proposed is the same instrument as that which 

 has been described, except that the same experimental bar C is al- 

 ways used, and that a springjs connected with the obhque threaded 

 screw d so as to cause the point of this screw to be always pressing 

 upon the end of the bar C. The temperature may be indicated by 

 means of a graduated plate connected with the screw with a vernier 

 attached analagous to those of a theodolite, by means of a system 

 of wheels like a watch ; or by means of a telescope mounted as be- 

 fore described. This thermometer labors under one objection, viz. 

 that the different metals do not expand with perfect uniformity as 

 their temperatures increase : but as in this instrum.ent no tempera- 

 tures much beyond the limits of atmospheric variation are proposed 

 to be measured, this objection vanishes, for, the metals of difficult 

 fusibility are said to expand infinitely near to uniformity by equal 

 increments of temperature within the range of atmospheric changes. 



Another application of the principles of the same instrument is 

 proposed for use, in connection with scales of equal parts, for ma- 

 king mathematical drawings. Every one who has attempted to 

 make very accurate mathematical drawings, must be aware that a 

 distance of any number of divisions of equal parts laid off at one 

 temperature, differs from that laid off from the same number at an- 

 other. It follows, that in accurately plotting the triangulations of 

 extensive surveys, a practical difficulty would be experienced, and 

 it is often experienced, when the plotting is continued through va- 

 rious atmospheric changes. The method proposed for obviating 

 this difficulty, is, to have several of the pairs of bars described as 

 maintaining a uniform distance between two points at different tem- 

 peratures, and let the distance MN, in them respectively, be one 

 inch, one foot, one yard, one metre, or any other convenient units. 

 By means of these units, the common brass scale of equal parts 

 usually seen in boxes of French mathematical instruments, may be 

 set each day when commencing to draw. If the temperature of the 

 room remain uniform during the day, and the scale be not handled 

 to increase its temperature, the difficulty arising from expansion so 

 far as it depends on the uniformity of the scales for measurement is 

 obviated. 



Those gentlemen engaged in plotting the triangulations of the 

 coast survey of the United States, under the direction of Mr. Hass- 

 ler, frequently experience the inconvenience arising from the expan- 

 sion of their scales and paper. I trust that the above suggestion 



Vol. XXX.— No. 2. 42 



