30 Transactions of the Royal Canadian Institute 



In the design of hot- water radiators now in general use, the ratio 

 of the heat suppHed by convection to that supplied by radiation varies 

 considerably with surrounding conditions. A thorough investi- 

 gation of this question would be of value. In the past it appears 

 that many of the improvements have been limited by economic 

 pressure merely to modifying existing systems progressively. 

 The systematic testing of a series of radical designs is required. 

 Theoretical considerations lead to the suggestion of many unfamiliar 

 types. 



A simple method of investigating the loss of heat from a radiator 

 in such a way that many of the usual difhculties are avoided is as 

 follows. Fill a small model of the radiator with cold water and fit 

 it with thermometers; place it in a large bath of hot water kept at 

 constant temperature and obtain a "warming up" curve; do the 

 same in a large enclosure of hot air kept at constant temperature; 

 place the model in a jar and do the same experiment before and 

 after evacuation of the jar; repeat the series with the temperature 

 conditions reversed and get "cooling" curves; examine also the effect 

 of increasing the natural convection in the enclosure. The analysis 

 of the "cooling" and "warming up" curves should lead to more 

 accurate results than are to be obtained from the application of the 

 available formulae and published data. The effect of the nature 

 of the surface on the emissivity could be studied at the same time. 

 Engineering tables give most conflicting data with regard to the 

 influence of the surface. 



Experiments of this character can be developed to check for- 

 mulae of general interest and importance. Information regarding 

 the temperature coefhciencts of radiation and convection could be 

 obtained, and the objective reality and the general characteristics 

 of the special films of air which are supposed to adhere to the sur- 

 faces of warming or cooling bodies, could be determined with the 

 aid of small thermocouples. The design of methods of procedure 

 which lead to the elimination of external losses from consideration, 

 should be the chief concern in such experiments. 

 (5) The influence of small atmospheric eddies and general turbulence 



on the validity of the theoretical and of the empirical formulae which 

 are applicable to convection problems, requires extensive exami- 

 nation. The interpretation of such instruments as the kata- 

 thermometer and the caleometer under turbulent open air condi- 

 tions is still in an unsatisfactory state, and it is of importance that 

 the data obtainable with such instruments which are so valuable 

 in the quieter conditions, should be ascertained more definitely 



