240 PROCEEDINGS OP THE AMERICAN ACADEMY 



single observation from the curves upon which the tabular values 

 were interpolated were: naphthaline 0°.l, benzophenone 0°.3, benzol 

 0°.06. 



6. The device used by us for controlling the pressure is easily ad- 

 justable, and sufficiently constant to afford, in connection with the 

 substances which we have investigated, a means of obtaining exactly 

 any desired temperature within the range measured. Thus this or a 

 similar apparatus may serve for testing mercurial thermometers at 

 several points, or for maintaining adjusted and known temperatures 

 for other purposes. 



Air Thermometer. 



Primary temperature measurements to be reducible to the absolute 

 scale (Thomson's) must be made by a gas thermometer, and the air 

 thermometer is the most available. The Jolly* form, in which the 

 closed and open arms of the manometer are connected by a flexible 

 tube, is the most convenient, and was adopted in this work. The dif- 

 ference of level of the mercury surfaces was, however, measured by 

 a special device. Vertically between the two arms of the mercury 

 column is placed a steel millimeter scale of 1.3 cm. square section, and 

 with straight edges. A T-square with double blade is held by the 

 hand firmly against the scale edge, and the blades, which project across 

 the face of the scale, pass, one behind, the other in front of the mercury 

 column. The square is then slid up or down until the plane of the 

 lower edges of the blades is tangent to the top of the meniscus, just as 

 in setting a barometer vernier. By placing the tubes so that there is 

 a bright light behind them, differences of level of the columns can be 

 read with errors of less than one tenth of a millimeter, the tenths being 

 estimated. The bulb used was about 15 cm. long, 4.5 era. in diameter, 

 and 0.5 to 1 mm. thick, with a capacity of about 200 cc, and the 

 volume of projecting stem was but 0.64 cc. The glass bulb and tube 

 were continuous over to the three-way cock to which the flexible tube 

 was attached. The class sauce tubes were about 1 cm. inside diameter. 

 The gauges, scale, etc., were carefully protected from heating, and their 

 temperatures obtained by suitably disposed thermometers. The bulb 

 was filled and emptied many times at 100°. The air was thoroughly 

 dried by calcic chloride, sulphuric acid, and phosphoric anhydride, and 

 its carbon dioxide was removed by sodic hydrate. Two determinations 

 of its coefficient of expansion made one before and one after the 



* Pogg. Ann., Jubelband. 



