Delicate Thermometers. 405 



by glass, prove, in a striking manner, the extent to which such 

 absorption occurs, and the persistence with which the absorbed 

 gases are retained. 



Both the delicate instruments ('39 and '616), in which the 

 difference of the readings was first observed, had had small 

 temporary bulbs attached for the purpose of calibration; and 

 it therefore seemed probable that the second opening of the 

 tube may have been the sole cause of the defects which they 

 exhibited. To settle this question, and to ascertain whether it 

 was possible to make instruments of such delicacy entirely 

 free from this defect, two other thermometers were manu- 

 factured, Nos '08 and '61. The delicacy of these was some- 

 what less than in the former instruments, owing to the 

 impossibility of procuring sufficiently fine tubes; the size of 

 the bulbs, however, was increased, that of '61 containing as 

 much as 46 grams of mercury. An estimation- figure, '05 

 mm., represented about 0*0005° 0. On examining these instru- 

 ments, in the same manner as previously, it was found that 

 they worked perfectly, the mercury registering exactly the 

 same temperature whether the column had risen or fallen to 

 the point of rest, equally satisfactory results being obtained 

 whatever portion of the stem was examined. Instruments of 

 this excessive delicacy are therefore perfectly workable; it is, 

 however, only by observing the utmost precaution in making 

 them that success can be obtained. The tube must on no account 

 be opened till the last minute, when the bulb is finished and 

 ready to be attached without a moment's delay; the bulb, as 

 soon as it is attached, must be warmed so as to fill the tube with 

 mercury and prevent the access of air through the upper end. 

 If any failure occurs in the attachment of the bulb at the first 

 trial, the stem must be rejected ; a second attempt would be 

 attended with the same results as putting on a second bulb 

 after the instrument had been made up. When once the 

 stem is filled with mercury, the tube may apparently be 

 opened several times at the top without damage being done, 

 and the bulb itself may be made 24 hours before it is attached 

 to the stem without being injured by exposure to air for that 

 time. 



It is a common practice of thermometer-makers to examine 

 the bore of a tube before it is made into a thermometer by 

 passing a thread of mercury along it, and often, indeed, the 

 stems are divided and fully calibrated before the bulb is 

 attached and the tube closed. From what has been ascertained 

 as to the effect of the air on the interior of the tube, it is obvious 

 that a tube which has been treated in such a manner will be 

 utterly useless for any really delicate instrument. 



