236 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 



Ultra-violet radiations of solar origin of wave length smaller than 

 2.950A. are absorbed by the atmosphere and do not reach us. There- 

 fore in order to obtain light which is truly bactericidal (of wave 

 length less than 2.800A.) we must have recourse to artificial sources. 



The quartz mercury-vapor lamp is the most powerful among these 

 latter. Luminescent mercury vapor is very rich in ultra-violet 

 light. Its ultra-violet spectrum reaches from X = 3. 650 A. to^ = 2.225A. 

 Quartz is transparent to all light of greater wave length than 1.500A., 

 and therefore to all of the rays of the spectrum given out by lumi- 

 nous mercury vapor. Since the ultra-violet rays of smaller wave 

 length than ^ = 2.80()A. are especially bactericidal, those between 

 2.800A. and 2.225A. render the quartz mercury-vapor lamp very 

 nocuous to all living cells, dangerous to anyone handling it without 

 precaution, as well as useful in the destruction of microbes. 



The quartz mercury-vapor lamp is fed by a continuous current. 

 It must be protected by suitable cooling devices. With proper pre- 

 cautions theoretically it will last indefinitely, practically a very long 

 time. The running of the lamp is very easily regulated either by a 

 "sentinel" lamp in the same circuit with the mercury lamp or by 

 simply noting the luminous state of the lamp itself. The lamp is 

 most readily lit by rocking it. As soon as a small thread of mercury 

 connects the two poles, a current of electricity is established and the 

 lamp becomes luminous. The lamp is then turned to its usual posi- 

 tion, the thread of mercury broken, but the current continues to pass 

 by means of the luminous vapors from the mercury. 



THE STERILIZATION OF DRINKING WATER BY MEANS OF THE QUARTZ 

 MERCURY-VAPOR LAMP. 



The most practical application of the bactericidal power of the 

 ultra-violet radiation from the quartz mercury-vapor lamp is for the 

 sterilization of drinking water. 



The difficulties met hi the sterilization of drinking water are well 

 known. Innumerable are the processes of filtration or of chemical 

 or physical purification which have been advocated and applied. 

 Some are of little value, others are effective but costly and requiring 

 too extensive areas of land and too much manipulation. For a long 

 while a method has been sought which is not only simple and eco- 

 nomical, but also absolute in its efficacy for the sterilization of drink- 

 ing water under such conditions as occur in the great majority of 

 cases where water is collected for a city's use. 



Just such a simple, sure, and economical process has been devised 

 by M. Nogier and myself. 1 This process makes use of the bac- 



1 J. Courmont et Th. Nogier, Sur la sterilisation de 1' eau potable au moyen de la lampe en quartz a 

 vapeur de mercure. C. R. Acad. des. Sciences, 22fevrier 1909. 



