4 Scientific Proceedings, Royal Dublin Society. 



the nozzle before it spread out, and weighing. These observations were repeated 

 from day to day in order to make sure that the sprayer was not varying in its action. 



Purity of the Water. 



In the previous work referred to, the oi-dinary distilled water of the laboratory 

 was used. We liad now access to water of a somewhat higher degree of purity. 

 Experiments were therefore carried out with this water, and also with samples of 

 different grades of purity prepared by mixing the distilled water in various pro-, 

 portions with tap-water (Vartry). In the course of each experiment samples were 

 drawn off from the spraying bottle, and the conductivity determined. The purest 

 water had a specific conductivity of 2'4 x 10"° ohnr^. 



Results : Charge on water from sprciyer driven at different pressures. 

 The continuous curves (fig. 1) show the total quantity of positive electricity 



33 30 33 40 



Fia. 1. 



per second communicated to the water when the sprayer is driven at various air- 

 pressures. Three curves are given corresponding to three samples of water of 

 different degrees of purity. The conductivity of each sample of water is indicated 

 on the curves. The dotted curve in the same figure shows the total cpiantity of 

 water per second issuing from the sprayer at different pressures. Combining this 

 curve with each of the other three in turn, we get the curves shown in fig. 2, 

 which give the charge per c.c. in electrostatic units corresponding to different 

 pressures on the sprayer. Some points may be at once noted. The charge per c.c. 

 depends on the purity of the water, increasing as the conductivity decreases. The 

 values are much higher than any previously reported, especially in the case of the 

 purer sample. Each curve shows a discontinuity ; the pressure at which the 



