136 PROCEEDINGS OF THE AMERICAN ACADEMY. 



and the mean of these locates with sufficient accuracy the maximum 

 point. 



So long as the shutter over the window of the containing box is 

 closed, the zero position should remain unchanged. With ordinary 

 jacketing, it was found that this was not the case. Indeed, in some 

 cases there was as much variation as 25°. This could be due to one 

 or all of three causes, viz. : convection currents, unequal heating of 

 the bulb or indeed any change of temperature of the apparatus, or to 

 small charges of electricity on the vane or containing vessel. 



To guard against the first of these, as has been said, the apparatus 

 was enclosed in a double-walled box with an asbestos interspace. But 

 since the McLeod gauge, which, of course, must be left uncovered, and 

 is connected with this instrument through various obstructed tubes, 

 responds quickly to any change of temperature in the room, any 

 change in this temperature will cause a flow of gas either towards the 

 transpiration instrument or away from it. To overcome this difficulty, 

 a house whose walls are of asbestos and double, with an air interspace, 

 was built surrounding the apparatus. In it were placed four electric 

 heaters, the current through which can be made and broken by a relay 

 worked by a battery circuit, in which is a thermostatic strip. In this 

 way the temperature of the whole space — about 24 cubic meters — 

 enclosed can be maintained very nearly constant. The temperature 

 was kept near 22° C. day and night, so that one might begin work 

 at any time and know that the apparatus was not far from a uniform 

 temperature. 



The second cause, that of temperature change in the instrument 

 itself, proved the most serious. Before the constant temperature room 

 was resorted to, invariably during the day with rising temperature 

 there was an advance in the zero position, i. e., the apparatus acted as 

 if the vane were illuminated, while during the night, as the tempera- 

 ture fell, the contrary was the case. The blackened surface would, of 

 course, absorb heat which was being conducted in through the walls 

 of the containing vessel faster than the clear surface, and a transpira- 

 tion would begin which would have the same direction as that arising 

 from the illumination of the vane. When the rise of temperature 

 ceased, then the two surfaces would gradually reach the same tempera- 

 ture, and in the absence of cause for transpiration the vane could be 

 put into its zero position without using force. Only by enclosing the 

 bulb of the apparatus in a pretty thick (1.5 mm.) brass shell (the object 

 of which is to distribute quickly the heat which enters from without) 

 and then packing the box full of cotton wool, leaving only a channel to 

 the mirror, E, closing the door of the box, and then packing it all around 



