244 Mr. Shelford Bid well on the 



Exp. 15. — One end of a brass tube was closed by a polished 

 disk of crystalline selenium, 11*2 millim. in diameter and 

 1*25 millim. thick, which was tightly fixed by an indiarubber 

 collar : to the other end was attached a glass tube which 

 terminated in an open point about 3 millim. in diameter. 

 The whole of the indiarubber and the edges of the Se 

 disk were thickly coated with shellac varnish. The closed 

 end of the tube was dipped into water and air was forced 

 into the open end. ~No bubbles were formed upon the Se : 

 if therefore Se be porous its pores must be very small. 



Exp. 16. — A transpiration experiment was then made. The 

 tube was fixed in an upright position with the Se disk at the 

 top and the glass point at the bottom, the latter dipping into 

 a basin of water. A wide-mouthed glass bottle was sup- 

 ported mouth downwards over the tube, enclosing its upper 

 end. Coal-gas was slowly admitted into the bottle from a 

 small nozzle. The indiarubber tube conveying the gas was 

 for seven feet (about 2 metres) of its length coiled up in a 

 vessel of cold water. When the gas was first admitted water 

 rose for a short distance inside the glass tube owing to the 

 cooling effect of the refrigerated gas : but after a minute or 

 two it began to descend again, and soon air-bubbles were 

 slowly but regularly given off at the pointed end. This con- 

 tinued until the bottle of gas was removed, when the bubbles 

 at once ceased and the water again began to rise in the tube. 



Crystalline Se is therefore sensibly porous, and no doubt 

 absorbs moisture from the air. 



§ 11. The Polarization of Selenium by Currents. 



Among the many remarkable properties of Se which were 

 first observed by Prof. W. Gr. Adams and Mr. R. E. Day in 

 their well-known research, is the power it possesses of setting 

 up a secondary or polarization current after being discon- 

 nected from a battery. As all who have worked much with Se 

 must know, the opposing and ever varying electromotive force 

 of polarization renders it very troublesome to make accurate 

 resistance tests by the ordinary bridge method. Unless it is 

 worth while to expend a considerable amount of time, one has 

 to be contented with two or at most three significant figures, 

 and even this degree of accuracy cannot be attained unless 

 care is taken to depress the battery and galvanometer keys 

 only for a moment *. 



* If the battery key be kept down for a few seconds, and after it is 

 raised the galvanon eter key be depressed, a considerable deflexion due 

 to the secondary current will generally be observed. For exact measure- 

 ments it is necessary to balance to a constantly changing false zero. 



