890 REPORT OF COMMISSIONER OF FISH AND FISHERIES. 



Test 1^0. 3, in rising temjyerature of air. * 



After test No. 1, the iustruments were thoroughly wiped and trans- 

 ferred to the cabin and hung side by side, bulbs on same line ; doors 

 and windows were closed and steam turned on to the heater. No. 3, 

 which had been hanging in cabin was included in the test. About 3 

 minutes were occupied in the transfer, during which they were in air. 

 TemDerature 64P. 



At ninety minutes the atmosphere of the cabin was so close that it 

 affected my accuracy in reading, and the experiment was therefore 

 ended. The No. 4 stood with the other standards in forty minutes at 

 7oA^, but then, as the temperature rose, passed them and showed about 

 2° difference in the balance of the readings. I can assign no cause for 

 this discrepancy, nor say which was the correct record, except that the 

 w^eight of evidence is in favor of No. 1, with which Nos. 2 and 3 

 closely agreed. Nos. 5, 6, 7, and 8 at the end were still below the 

 true temperature. This I attribute to the fact that, probably, we had 

 not perfectly dried all crevices in the metal cases, and that we got the 

 evaporating point in lieu of the true temperature. At the end of fifty 

 minutes the cabin door was*opened for a moment; this produced an in- 

 stant change, which Nos. 1, 2, 3, and 5 recorded fully. No. 4 fell a 

 little, but the three Casellas seemed to make no change on account 

 of it. 



