1 88 CARNEGIE INSTITUTION 



which is convoluted between the double walls of the house and pro- 

 vided with drips. The tube enters the inner chamber near the ceil- 

 ing and passes round the entire inner space, always at a slight slope, 

 so that all condensed moisture trickles backward. From the air tube 

 air escapes into the instrument room through small holes in the side 

 of the tube. The result is, for some purposes, a very satisfactory 

 one, attained almost without expense. In a larger laboratory a 

 somewhat different method would probably be more convenient and 

 more effective. 



Difficulties of Varying Temperatures. 



Attempts have been made in Europe to control the temperature 

 of large apartments by providing them with double metallic walls 

 in which hot or cold solutions circulate, but these efforts have not 

 been successful. In warm weather, when cold solutions must be 

 employed, the walls drip with moisture, the instruments suffer, and 

 the operators fall ill. This would be avoided by supplying the apart- 

 ment with air not merely cooled but dried, just as a room in winter 

 may be heated by a hot-air furnace. 



In this country more attention has been given than in Europe to 

 cooling apartments with dry air. The Bureau of Standards has de- 

 veloped arrangements for this purpose which will be in operation 

 in a few weeks, and it is said that the Stock Exchange in New 

 York is being similarly equipped. The experience obtained by the 

 Bureau of Standards should be carefully considered before any specific 

 plan is adopted for a geophysical laboratory. 



Importance of Uniform Temperatures. 



In my opinion, a modern laboratory should be supplied in summer 

 with dry, cool air, the temperature of which is under control. Such 

 air of appropriate temperature should be admitted to the cold sub- 

 terranean chambers when required, and should be furnished to the 

 ordinary laboratories in such quantity as to keep their temperature 

 down to 20° C. = 68° F. In winter the rooms must, as a matter 

 of course, be warmed. If the problem of maintaining a laboratory 

 at constant temperature is not wholly simple, it is surely of small 

 complexity as compared with those of physical research, and it can- 

 not be doubted that were its solution requisite to the success of a 

 commercial enterprise, an appropriate method would soon be de- 

 veloped. Yet it is unquestionable that physical research would 



