CASELLA-MILLER DEEP-SEA THERMOMETER. 899 



portioD of bulbs. This induced quicker action, but still slower tbau a 

 bath of 350 ; and at last, as iu former cases, the mercury apparently 

 ceased to fall at 35^. 



This experiment is put in simply for the purpose of guarding other 

 experimenters against the mistake. 

 The standard, as in last year's tests, did not vary from 32°.05. 

 Kespectfully, 



L. A. BEAIIDSLEE, 

 Commander, United States Navif. 



