May, 1910.] Proceedings of the Asiatic Society of Bengal. 



cxi 



Dr. Leather, Pusa, holds similar views. He says : " It is 

 easier to maintain a vessel at a temperature greater than the 

 atmospheric one than below it. It is better to select a tern- 

 perature near to the laboratory maximum so as to provide 

 principally for hot- weather conditions." 



A temperature of 15° is above the mean temperature of 

 England and Central Europe, and if what might be termed a 

 corresponding temperature were fixed for India, that tempera- 

 ture should be above 28°. Since chemists and physicists have 

 sometimes found 15° to be an inconveniently low temperature 

 for a standard in the West, it would seem that 30° would not be 

 found to be an inconveniently high temperature for a standard 

 of reference in India and the Tropics. 



Mr. R. LI. Jones, Madras, advocates 28° as a standard in 

 electrical w T ork. He remarks : — 



' b It appears to me that 28°C. is the most convenient tem- 

 perature to adopt as a standard. The reasons why this value 

 is taken are briefly the following : — 



4 c For accurate standardising and high precision work it is 

 necessary to carry out the operations in a constant temperature 

 room. It is also very desirable to secure this constant tem- 

 perature without artificial means of heating and cooling, if 

 possible. Hence in selecting a standard temperature it is ad- 

 visable to fix on such a temperature as can be secured with ease 

 and certainty and kept constant with the least amount of help 

 from subsidiary machinery. 



" Constant temperature rooms are generally built under- 

 ground in the middle of a massive block of buildings so that 

 the daily variations in temperature are quite inappreciable. 

 The only variations in temperature in such a room are the 

 seasonal or annual changes. The further down the room is the 

 less will these changes be — at 20 or 30 feet they arc very small. 

 The temperature at such depths is in India close to 82°F. or 

 28°C. according to the best observations. This is also approxi- 

 mately the mean temperature of the sea in the tropics and of 

 India as a whole, I believe. Hence it seems natural to take 

 28° as the standard temperature as it appears to be the mean 

 temperature of the equatorial belt of the globe. The fluctua- 

 tions from it in ordinary work will be more evenly distributed 

 on either side of it, and their average magnitude (irrespective of 

 sign) will consequently be less than if any other temperature 

 were taken as standard. The corrections to be applied will 

 hence be smaller and the uncertainty on this account will he 



reduced." 



These observations are of interest, but we need hardly 

 consider sea water, and our experience is that the majority of 

 laboratories in India are not constructed of massive materials, 

 nor are the usual operations conducted in underground cellars. 



