Calorimetric Thermometers. 341 



lerse 



5> 



id after being heated 10°. 

 „ cooled 10 ., 

 „ heated 10 ., 

 „ cooled 10 ., 

 „ heated 10 .. 

 ,, cooled 10 .. 

 „ heated 10 .. 



II 



Registered. 



.. 13-985 

 .. 13-95 

 .. 14-015 

 .. 13-91 

 ,. 14043 

 .. 13-955 

 . 14045 



Theory, 

 o 



14-00 



13-93 



14029 



13-933 



14-044 



Mean .... 



Difference, 



-•05 



+•085 

 — •119 



55 





+ •110 

 -•089 



" 









-091 



Thus, contrary to expectation, this instrument not only 

 shows a difference in reading when the column is rising or 

 falling to any temperature, but that difference is even greater, 

 and considerably so, than with No. 62839. This is certainly 

 remarkable, and tends to show that a full explanation of this 

 difference has not yet been obtained, for the double bulb 

 would undoubtedly possess a greater rigidity than the single 

 bulb of the other instrument ; although the combined length 

 of the two bulbs is 110 millim. as opposed to 72 millim., the 

 diameter of the tubing of which they were made was con- 

 siderably less, 4*83 millim. instead of 6*86 millim. (internal 

 measurements), and the wall of the tubing also was consider- 

 ably stouter, 0'76 millim. instead of 0'635 millim., and a 

 small improvement in these last two items would cause a very 

 considerable increase in the rigidity. Another peculiarity in 

 this instrument is that the difference in the readings is inde- 

 pendent of the height of the column of mercury; or at any 

 rate very nearly so. 



Series I. of the above experiments affords good proof that 

 the difference in readings is independent of the distance 

 through which the column has fallen or risen; the mean dif- 

 ference of the first three numbers in the last column being 

 •098, and that of the last three *093; numbers practically 

 identical, although the fall or rise was 1 degree in the former, 

 and 10 degrees in the latter. 



Although this difference, due to what I have termed the 

 permanent alteration in the size of the bulb, is greater with 

 63616 than with 62839, it is otherwise with the temporary 

 alteration. An experiment similar to the above, but the tap- 

 ping of the instrument being omitted, gave a difference of 

 0°*23 in the mercury standing at 45*8, whereas with 62839 it 

 amounted to 0°'6 (at 27°) ; these numbers representing the sum 

 of the permanent and temporary alterations, the latter only 

 will be represented by (-23- -096 = ) 0°"13, and (;6--027=) 

 0°*57 in the two instruments respectively. This tends to con- 

 firm an opinion expressed by Mr. Casella, that a considerable 



