467 



diately established, and continued so long as these conditions are 

 maintained. 



The experiments were made in a room 12 feet square by 8 feet 

 6 inches high ; the windows and chimney being carefully secured, and 

 all crevices closed, by pasting paper over them, the floor carpeted, 

 the door double, and the inner door surrounded with list. The 

 outer wall, having a north aspect, was so sheltered by surrounding 

 buildings that the direct rays of the sun never fell upon the window. 

 Discs of delicate tissue-paper were suspended in several parts of the 

 room, to indicate currents of air, if any existed, and observations 

 were taken only when these were perfectly quiescent. 



Mason's hygrometer was first employed in these experiments, to 

 test the presence of a current of air in the tube ; on the principle 

 that as evaporation produces cold, and as evaporation is increased by 

 a current of air, the wet-bulb thermometer would show a greater de- 

 pression if any current existed, than if the air were perfectly quiescent 

 within the tube. The tube was placed in the middle of the room, 

 and isolated from the floor by a cylinder of thick glass laid under it. 



It was found that in ninety- one observations of the hygrometer, 

 suspended in the free air of the room, the mean depression of the 

 wet-bulb thermometer was 3'9 Fahr., while in ninety corresponding 

 observations, with the hygrometer at the lower aperture of the tube, 

 the mean depression was increased to 4*9 Fahr., clearly indicating 

 the existence of a current of air within the tube. 



Partial closure of the upper orifice of the tube, by placing a piece 

 of fine muslin upon it, produced a sensible influence on the hygro- 

 meter. In seventeen observations with the tube thus partially ob- 

 structed, the mean depression was 2 0< 5 Fahr. ; but in an equal num- 

 ber of comparative observations, with the tube perfectly free, the 

 mean depression was increased to 3*12 Fahr. ; showing a consider- 

 able diminution of the force of the current within the tube, as a 

 result of the partial obstruction of its upper aperture. 



Similar comparative observations, with the hygrometer placed at 

 the upper aperture of the tube, yielded similar results. 



In these experiments the lower extremity of the vertical tube was 

 bent thrice at right angles*, for convenience in making the obser- 



* The tubes used in these experiments were bent either at their lower or upper 

 extremities for convenience merely. 



