74 Notes concerning the Phenomena 



as in the case of the dry air, and, consequently, the force 

 causing recession being weaker would enable the balance to 

 swing nearer the name. 



" In considering, in the next place, the effect of con- 

 densation as a cause of approach, it was thought that if a 

 balance could be placed under such conditions that rapid 

 condensation or absorption would ensue under the influence 

 of heat, the effect might be so exaggerated as to be made 

 very conspicuous. For this purpose the property that oil of 

 vitriol possesses of rapidly absorbing moisture was taken 

 advantage of. 



" Experiment 7. — A balance was constructed of a platinum 

 stem, with small plates of pumicestone at the ends, and 

 suspended horizontally within a flask by a fibre of silk. So 

 long as the pumicestone was dry the balance swung, but 

 slowly, towards the flame ; when, however, it was moistened 

 with oil of vitriol the motion was greatly accelerated. In 

 addition to this increased rate of motion towards the flame, 

 the following interesting behaviour was observed. The 

 heavier end always came towards the flame, even when the 

 flame was placed opposite the lighter. It then rapidly still 

 further increased in weight, by the attraction of moisture, 

 until sufficient had accumulated to form a drop. Directly 

 this drop fell, the end next the flame rose, being then the 

 lighter, and forthwith commenced to swing round until it 

 had changed positions with the other end, which, in its turn, 

 increased rapidly in weight, let fall a drop, and immediately 

 thereupon began to swing round as before. This periodic 

 exchange of positions continued so long as the drops 

 continued to fall. 



" The question as to why the lower end of the balance 

 should always approach the flame, now offered itself for 

 solution. It was conceived that it might depend upon the 

 relative amount of condensation that was taking place on 

 the two ends ; or again, that it might be a direct result of 

 the action of convection currents. In the endeavour to 

 settle this point a number of experiments were instituted. 



" Experiment 8. — A. very delicate balance, consisting of a 

 fine glass stem carrying small pith-discs at its extremities, was 

 suspended horizontally within a moist flask. A flame was 

 placed near and the movement of the balance attentively 

 watched, when it was found that its behaviour corresponded 

 with that of the balance in the last experiment. The end 

 next the flame was observed to rise slowly ; on its attaining 



