REMARKABLE CONTRACTION AND EXPANSION. 197 



let any gas escape. Restore it to its position, and read off on the scale. It is then 

 ready for a second trial. 



890. The difference between continuous and interrupted observation is this, that in 

 the latter we pause to wash out the muriatic acid ; and though this is effected by the 

 simplest of all possible methods, continuous observations are always to be preferred 

 when they can be obtained. 



891. I have extended this paper to so great a length that many points on which re- 

 marks might have been made must be passed over. It is scarcely necessary to say that 

 the sentient tube must be uniformly and perfectly clean. As a general rule, also, the 

 first observation may be cast aside, for reasons which I will give hereafter. Farther, it 

 is to be remarked, as it is an essential principle that during different changes of volume 

 of the gas its exposed surface must never vary in extent, the liquid is not to be suffered 

 to rise above the blackened portion at d. If the measures of the different parts be such 

 as have been here given, this cannot take place, for the liquid will fall below the fiftieth 

 division before its other extremity rises above d. 



892. The same original volume of gas in a d will last for a long time, as we keep 

 replenishing it as often as the fiftieth division is reached. 



893. The experimenter cannot help remarking, that on suddenly exposing the sen- 

 tient tube to a bright light, the liquid for an instant rises on a scale, and on dropping the 

 cap, in an instant falls. This important phenomenon, which is strikingly seen under 

 the action of an electric spark, I shall consider hereafter. 



894. In conclusion, as to comparing the tithonometric indications at different times, 

 if the gases have the same constitution, the observations will compare ; and if they have 

 not, the value can from time to time be ascertained by exposure to a lamp of constant 

 intensity. To this method I commonly resort. 



895. From the space occupied in this description the reader might be disposed to 

 infer that the tithonometer is a very complicated instrument, and difficult to use. He 

 would form, however, an erroneous opinion. The preliminary adjustment can be made 

 in five minutes, and with it an extensive series of measures obtained. These long 

 details have been entered into that the theory of the instrument may be known, and 

 optical artists construct it without difficulty. Though surprisingly sensitive to the ac- 

 tion of the indigo ray, it is as manageable by a careful experimenter as a common dif- 

 ferential thermometer. 



