Effect of Pressure on Thermometer-bulbs. 409 



in a manner analogous to that often noticed with thin tin 

 plate vessels, where a small addition or removal of pressure 

 will cause a sudden and considerable alteration in the form of 

 the vessel. It would certainly be advisable, in the case of 

 any thermometer required for very delicate work, to examine 

 it under pressure to ascertain whether its action be uniform 

 or not. In order to reduce the chances of irregular action, 

 it is necessary to render the bulb as inexpansible as possible. 

 From the fact that the thermometer 63616 showed a slight 

 amount of irregularity in its action, we may place the limits 

 of expansibility desirable in a very fine instrument at a number 

 between that of this instrument and that of 65108 ; the co- 

 efficient of expansion should not exceed 0' 000,000,03, or the 

 apparent expansion 0'02 millim. of the mercurial column per 

 millim. of pressure. 



An examination of table ii. of the preceding communica- 

 tion will show at once the increase of rigidity obtained by 

 having the bulb made out of a glass cylinder instead of being 

 blown before the lamp ; the instruments '83 and '39 B are the 

 only ones mentioned in this table which had blown bulbs, 

 and the coefficients of expansion in their cases are higher 

 than in any other case, although their bulbs were only J and 

 ^ as big as those of most of the other instruments. Again, the 

 thermometers '83 and '16 were identical in all respects except 

 as regards their bulbs, and here the blown bulb ('83) possesses 

 only half the strength of that made out of cylinder. 



Although it seems probable, prima facie, that a blown bulb, 

 however well constructed, would not be so uniform as a 

 cylinder bulb, these facts of course do not prove that such is 

 necessarily the case, as the thickness of the walls of the bulbs 

 was not known ; but it does prove that, by ordering a thermo- 

 meter with a cylinder bulb, we should in all probability get an 

 instrument possessing nearly twice the strength of one with a 

 blown bulb. A further very considerable addition of strength 

 may be gained by having the bulb made double instead of 

 single. The instruments Nos. '616, '08, and '61 possessed 

 double bulbs made out of glass cylinder ; and a comparison of 

 them with '83 B and '39, which had single bulbs made out of 

 cylinder also, will show the advantages of the double bulb. 

 For thermometers to be used in a liquid which is stirred in a 

 thoroughly efficient manner, the thickness of the walls of the 

 tube forming the bulb may be very considerable. The instru- 

 ments Nos. '08 and '61, which contain between 40 and 

 50 grams of mercury, will take the temperature of the liquid 

 in the calorimeter in about 5 seconds, although their bulbs 

 are 0*75 millim. thick in the walls. 



