PROGRESS OF METEOROLOGY IN 1889. 



219 



made by Tonuelot of hard glass whose cbemical composition is specified. 

 Four thermometers were compared having scales from — 5° C. to 104°, 

 and a length of 70'='", the length of a scale-division being 5.7""" ; and 

 fonr other mercurial t liermometers 54*="' long for low temperatures, with 

 scales from —32° to 4-39°, to which an extension is added which includes 

 scale readings from 95° to 103°. Between these eight mercurial ther- 

 mometers there were no systematic differences, and they agreed with one 

 another very well, the greatest difference in an observation being 0°.006. 

 The gas thermometer is constructed so that the portion that is not 

 brought to the temperature of the bath is extremely small. To attain 

 this a large cylinder of iridio-platinum (1.10"' high and 3G""" diameter) 

 is used as a thermometer box. It is a gas thermometer of constant vol- 

 ume. The pressure at 0° is about T" of mercury. The description of 

 the manometer and the detail of the comparisons will be found in the 

 Archives des Sciences phys et nat. Bd. xx, 1888. Tlie following is an 

 abstract of the differences found between a mercurial thermometer 

 and the three gas thermometers containing hydrogen, nitrogen, and car- 

 bonic acid, whose fixed points agreed Avith the mercurial. 



Differential barometer. — Director Criils, of the Rio Janeiro observa- 

 tory, has invented a new form of differential barometer for use in hyp- 

 sometry. Each arm of a JJ -shaped tube terminates in an air-tight box 

 of known capacity provided with a stop-cock. If the \J be jiartially 

 filled with a suitable liquid and both cocks are opened, the height of 

 the liquid in the two arms will be the same. If now the cocks be 

 closed and the instrument be taken to a station of different elevation 

 and only one of the cocks be opened, the closed box will retain the air- 

 pressure of the first and the open end will assume that of the second 

 station. The level of the liquid will now be different in the two arms 

 by a height which will depend on the difference of j)ressure at the two 

 stations. 



A light liquid like water can be employed, and yet the instrument 

 may be ke[)t of convenient size. If the stop-cocks are small, they may 

 be opened alternately, each just long enough to partially equalize the 

 difference of pressure in the boxes ; with each opening the level of the 

 ends of the licpiid columns is to be read otf, and the sum of the differences 

 measures the difference of pressure. {Am. Meteor. Journal, vi, p. 181.) 



