166 



Mr. Nixon on the Theory of the 



[Sept; 



n 



dividing the product by 480, add or subtract the quotient accord- 

 ing as the mean temperature is above or below 32°. 



In order to introduce the correction for the diminished speci- 

 fic gravity of the quicksilver in the vertical line, we must 

 augment the constant coefficient g^, or to 60246, and call the 



dilatation of air ^^ per degree. 



The indices of the logarithmic differences being sometimes 

 negative as well as affirmative, it will be advisable to disregard 

 them altogether, considering the tabular logarithms of the pres- 

 sures as decimals. 



Construction of the Barometer, 



Tiie syphon barometer consists of a glass tube bent in the form 

 of an inverted syphon, filled with very pure mer- 

 cury freed from air by carefully boiling it when 

 in the tube.* Each branch being furnished with 

 a scale of inches having their zeros coinciding in 

 level, the difference of the observed heights of 

 the summits of the (perpendicular) mercurial ^ 

 columns will be equal to the height of the baro- 

 meter or pressure of the atmosphere. In order 

 to dispense with the shorter scale, the index of 

 the longer branch is so constructed as to slide 

 down to the level of, and extend horizontally to 

 the summit of the mercury in the other branch. 

 In some barometers the zero of the scale is placed 

 anywhere above the shorter leg as at «, and the 

 inches are numbered upwards and downwards 

 so that the su7n of the two measurements, instead of their differ- 

 ence, exhibits the pressure. An augmentation in the pressure 

 of the atmosphere having taken place, a depression of the mer- 

 cury in the snorter branch, and an equal elevation in the longer 

 one, of half the quantity of the variation restore the equilibrium 

 of pressure. To obviate the trouble of measuring the difference 

 of level of the two columns, the shorter one only has been pro- 

 vided with a scale, having the half inches numbered as whole 

 inches ; — a method which renders it impossible to make a proper 

 correction for the variation of temperature of the mercury. (We 

 might inquire why the syphon itself, being laterally confined 

 within cylindrical rings, might not be raised or depressed by 

 means of a screw fixed below y, so that the summit of the 

 shorter column might always comcide in level with the zero of 

 the scale fixed at a; . - 



The following methods have been resorted to with a view to 

 render the instrument portable. 1. The two branches, being 



* It has been proposed as an improvement to dll the tnbe in a vacunm. 





