898 EEPORT OF COMMISSIONER OF FISH AND FISHERIES. 



nometer, and in a few moments the column dropped away from the 

 sphere. This occurred twice, when I concluded that it must cool more 

 slowly, so left it in the boiling water, to cool with it. This produced the 

 desired result, and the hot side became perfectly restored. On the cold 

 side I substituted a mixture of ice and salt, in which I placed the bulb, 

 and with an ordinary cologne spray-bottle, substituting sulphuric ether, 

 threw a spray on the bulb, which acted as did the steam. I was com- 

 pelled to leave the instrument to warm gradually in a vessel of ice- 

 water, and in an hour had it in good working order. 



I send herewith a set of diagrams of the curves produced by the dif- 

 ferent baths tabulated, and am, 

 Very respectfully, 



L. A. BEARDSLEE, 

 Commander, United States Navy. 

 Prof. S. F. Baird, 



United States Fish Commissioner. 



To ascertain the correctness of the standard used in the foregoing ex- 

 periments, it was tested in the apparatus provided by the Signal Corps 

 for testing thermometers — simply a tin can with a perforated tin false 

 bottom about midway of its height to permit the leakage from the melted 

 ice to escape. 



A thermometer manufactured by Tagliabue for the Commission, with 

 cylindrical bulb 1^ inches long and y\ in diameter, graduated from 30°' 

 to 140°, on a scale of 8f inches, was compared at the same time. The; 

 bulbs were immersed and packed snug in pounded ice. The Tagliabue. 

 recorded Sl^.G, and the standard 32o.05. The experiment was repeated 

 with the same result. 

 Eespectfully, 



L. A. BEARDSLEE, 



Cotmnander. 



Wood's Holl, September 23, 1875. 



An additional set of tests were made with the Cassella-Miller ther- 

 mometer in use this summer at Wood's Hole (No. 1,844). They were, 

 made with the utmost accuracy possible, with the same standard in use 

 last summer, and the aquarium. They have developed the reason of 

 the anomalous action of the thermometers in ice. 



In test No. 1, in ice, the instrument was placed on its back in a trench 

 cut in the ice, and the bulbs covered with pounded ice. The mercury 

 corresponded in its action very closely to the action of a bath of 50<3 

 until it reached that point. 



In test No. 2 the upper portion of the ebonite guard was removed, 

 thus letting the crushed ice come into immediate contact with the uppen 



