75 



By far the most satisfactory method is to fill in between the walls of the tin 

 box with mineral wool, and to use wire coils and an electric current to heat the 

 enclosure. 



In the earlier part of my work I used distilled water from the Chemical 

 Laboratory. Subsequent tests showed that it contained considerable organic 

 matter. I am now using water which has l)een distilled three times in glass; once 

 with permanganate of potassium to remove organic matter. My observations 

 range from 0° to 80°, and cover a period of four months. 



Briefly, my conclusions are as follows: 



Between 0° and 80° the temperature coeflScient curve is concave toward the 

 X axis, when we use tensions as ordinates and temperatures as abscissas. This 

 coefficient increases with the temperature, its value being about .17 dynes. 



The formula usually used to represent the tension (T) at any tempera- 

 ture vt°j is 



Tt°=To— .14 t°. 

 I find that the tension can not be expressed as a linear equation, and that 

 .14 dynes is too low for the average temperature coefficient. 



Much of my work so far has been toward perfecting the method and my 

 apparatus. I am now making some observations, using exceedingly thin 

 mica frames, and standardized thermometers reading to one one-hundredth of a 

 degree. For temperatures below 0° I shall use the method described by Messrs. 

 Humphreys and Mohler in the " PhysicaUEeview," March-April, 1895. I shall 

 endeavor to extend my observations above 100° by using the capillary tube 

 method, the water and tubes being enclosed in an air-tight plate glass box and 

 under whatever pressure is necessary to maintain the desired temperature without 

 boiling the water. 



Physical Laboratory, Indiana University, December, 1895. 



Strains in Steam Machinery. By W. F. M. Goss. 

 Masses of metal when of considerable strength and weight would ajjpear to 

 be proof against distortion under the influence of any force which may be brought 

 to bear upon them. We think of the strength of metals, but it is not often that 

 we consider their elastic property, yet, physically speaking, nothing, probably, is 

 more elastic than steel. A piano wire, if tightly strung, increases its length, and 

 if loosened again it contracts. Within certain limits it behaves precisely like a 

 spring. When force is ajjplied it stretches, and when the force is withdrawn it 



