144 



rature, as affecting the results given by the instrument, were shown 

 to be effectual without the aid of the thermometer. 



The difficulties to be contended with in taking accurate observa- 

 tions were shown to be local variation, diurnal variation, and some 

 irregularity in the action of the mechanical parts of the instrument 

 itself. These difficulties were examined seriatim, and modes of ap- 

 proaching to their corrections were explained. The modes of com- 

 pensation for variations of temperature affecting the instrument were 

 shown as at present practised by the makers : the diaphragm-box 

 being compensated by means of the introduction of a small portion 

 of aeriform fluid, instead of being allowed to act with a perfect va- 

 cuum, and the metallic connexions between the diaphragm-box and 

 the index being compensated by compound arms or connexions of 

 steel and brass so adjusted as to neutralize mutually the respective 

 contraction or expansion of each at variations extending to 100 de- 

 grees of temperature. 



The mode now practised by makers of graduating the aneroid 

 (when thus compensated) by comparison with a standard mercurial 

 barometer, was stated, and it was suggested that improvement on this 

 practice might be made by reference to standard elevations running 

 up to 2000 feet at least in Great Britain. Practical examples were 

 given of the use of the instrument in Ceylon, showing the variations 

 of the aneroid (when properly checked) to lie between 1 foot and 6 

 feet, as compared with the surveyor's spirit-level : other examples 

 were given of practice on the Great Western, South-Eastern, and 

 North Kent Railways, varying from the true levels from 6 inches 

 to 6 feet, over distances of between 300 and 400 miles. 



The paper concluded with Tables in the Appendix, and with dia- 

 grams explanatory of the construction of the instrument ; the Tables 

 being intended to illustrate the effects of diurnal and also of local 

 variation within the tropics (in Ceylon), and also in England. 



