THERMOMETERS.] 



METEOROLOGY. 



1131 



instrument-maker ; on the contrary, they are intended to 

 make known to him the defects which he should look for 

 in a thermometer, and which, if discovered, should 

 cause the thermometer to be rejected. 



The Tube mimt be Equal in Bore throughout. If the 

 bore or diameter of a thermometer tube be not equal 

 throughout, it is evident that the amount of linear ex- 

 pansion cannot be equal, and that the instrument will 

 be absolutely worthless. A very easy means of gauging ! 

 this equality is the following : Having selected a piece 

 of thermometer tube open at both ends, tie on to one ex- 

 tremity a riyid bottle of india-rubber, and dip the other 

 end in mercury, thus (Fig. 22). Pressing the bottle, a 

 portion of the atmospheric air will be expelled ; then, 

 Fig. 22. allowing the bottle to ex- 



pand, some mercury will 

 enter, forming a mercu- 

 rial column, the length 

 of which admits of being 

 measured. Let it be 

 measured by means of a 

 pair of compasses ; and 

 let the mercury be driven 

 to various parts of the 

 tube, and the measure- 

 ments repeated. If the 

 tube be of equal bore 

 throughout, the mercu- 

 rial column will necessa- 

 rily be of equal length 

 throughout ; and if the 

 tube stand this test, it 

 maybe considered good. 

 Instead of the india- 

 rubber bottle the breath 

 might be employed, but 

 it would be attended 

 with the disadvantage of 

 moistening tbe tube. However, it is possible to use the 

 breath without prejudice, if the operator take the pre- 

 caution of blowing through some material absorptive of 

 moisture. The next step in the manufacture of a ther- 

 mometer consists in fusing one end of the tube, and 

 blowing it into a bulb ; this, a;;ain, should be effected by 

 means of the caoutchouc bottle, lest moisture be depo- 

 sited. Mercury has now to be introduced, which is ac- 

 complished as follows : 



The thermometer tube having been bent as represented 

 (Fig. 23)', its open extre- Fig 23. 



mity is immersed in a 

 vessel of mercury. Heat 

 being now applied to the 

 bulb, the air therein con- 

 tained is expanded ; and 

 the heat being removed, a 

 partial vacuum results, to 

 fill which, mercury rushes 

 in. By repeating the ope- 

 ration, the mercury al- 

 ready contained in the 

 bulb is vaporised, and the 

 vapour, expanding, drives 

 out all the remaining at- 

 mospheric air ; so that on 

 the removal of heat, the 

 whole tube, bulb, stem 

 and all becomes filled 

 with mercury. The bent 

 part of the tube is now 

 broken off, and the final 

 quantity of mercury duly 

 apportioned to the tube. 

 The amount of this apportionment can be only deter- 

 mined by practice ; but, in general terms, it may be de- 

 scribed as being such a quantity that, at the boiling 

 point of mercury, it shall nearly extend to the extremity 

 of the tube.* 



Care i required J In , thU proc, that the cold metal .hall not reddenly 

 Impinge on the hot bulb, u that would certainly break it.-En. 



The next process is one of extreme delicacy. It has 

 for its object the sealing or melting the open extremity 

 of the thermometer tube, without admitting the slightest 

 portion of atmospheric air, the presence of which would 

 materially interfere with the delicacy of tbe instrument. 

 The operation is conducted as follows : The open end of 

 the tube having been melted in the blow-pipe flame, is 

 drawn out to a fine termination, thereby diminishing 

 still further the internal bore of the tube, and rendering 

 the final closing of its orifice more easy. The mercury 

 in the bulb is heated once more ; and at the very instant 



Fig. 21. 



when it is seen to fill the tube and to be in the act of 

 overflowing the capillary extremity, the tube is closed by 

 the fine jet of blow-pipe flame b (Fig. 24). It only now 

 remains to graduate the thermometer. 



Graduation. The instrument just described is, in its 

 present condition, a thermoscope, or indicator of heat ; 

 but it is by no means a heat-measurer to convert it into 

 which, it requires to be graduated into a determinate 

 number of equal parts. These parts are usually called 

 degrees ; but it must be remembered that the so-called 

 degrees have no ratio whatever to any particular unit. 

 They are not like the degrees on the circumference of a 

 circle, each one of which is related to the trigonometrical 

 ratio, or the ratio of the circumference to the diameter. 

 Thermometric degrees are altogether arbitrary, except in- 

 somuch as certain usages in this respect obtain. This 

 much is, however, invariable and universal the divi- 

 sional parts or degrees must be established between cer- 

 tain limits, fixed by nature. The limits usually adopted 

 in the manufacture of thermometers are the boiling and 

 the freezing of water, which phenomena always, under 

 similar conditions, take place at similar respective tem- 

 peratures. Founded on the bases of these limits, three 

 principal scales of graduation have been devised. They 

 are the centigrade scale, or scale of Celsius ; the scale of 

 Reaumur, and the scale of Fahrenheit. The latter is 

 mostly used in this country ; the former is chiefly 

 adopted on the continent. 



Let us commence our illustrations with the centigrade 

 scale, as being most easy. The term centigrade seems to 

 be significant of a hundred divisions. In point of fact, 

 the centigrade scale has for its 0, or zero, the temperature 

 at which water freezes, and, for its 100, the temperature 

 at which water boils ; the intermediate space being 

 divided into 100 equal parts. And if it be desired to 

 carry the graduations above or below the gauge limits, 

 this is accomplished by measuring off equal parts by 

 means of a pair of compasses. 



Reaumur's scale has also its zero point, at the elevation 

 of mercurial column corresponding with the freezing 

 point of water ; but at the other extremity of its scale 

 the boiling point of water is considered to indicate 80 ; 

 hence the intermediate space is divided into 80 equal 

 parts. The following diagram (Fig. 25) will reuihr 

 evident the peculiarities of the ordinarily employed 

 thermometrio scales. 



Conversion of One System of Graduation to Another. 

 This conversion of thermometric scales is frequently 

 necessary. The rules for effecting the conversion are 

 evident on reflection ; nevertheless it is well to reduce 

 them to general formulae. If all these scales counted 

 their zero from the same point, the method of convert- 

 ing one scheme of graduation into another would be still 

 easier than we find it. Actually, some little confusion 

 at first arises from the circumstance that Fahrenheit's 



