156 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1941 



vantages of low-cost installation and ability to store heat at a higher 

 temperature would be offset by the disadvantage of lower efficiency 

 of collection is a point requiring study. 



Wliether the use of flat-plate collectors together with a storage 

 system is economically possible for house heating or air conditioning 

 in certain areas of the earth, whether other types of collector will 

 prove cheaper for these uses or for power generation, whether power 

 generation from solar heat demands the development of a new 

 heat-engine cycle, and whether power generation by any process 

 dependent on direct conversion of sunlight into heat with consequent 

 unavoidable losses due to the degradation of energy is sound — these 

 are questions which it is hoped this program will help to answer. 

 Regardless of the result, the present considerable and increasing 

 importance of solar heat for hot water in certain parts of this coun- 

 try indicates the need for a comprehensive study of the factors 

 involved in the design of collectors. 



Now to come to a second project, related to the one just dis- 

 cussed. Conventional heat-power plants are characterized by a cost 

 of power production depending enormously on the capacity of the 

 plant; and we have seen that solar power does not now look very 

 attractive when compared to large-scale operation of steam plants. 

 If, on the other hand, it were possible to operate small solar plants 

 with an efficiency comparable to large ones, the comparison with 

 fuel-fired plants might lead to some very different conclusions. 

 So far as the collectors of the sunlight are concerned there is little 

 indication that the cost should be other than proportional to the 

 amount of collector area. If then it were possible to devise an 

 engine with moderate efficiency even in small units, one might have 

 something worthwhile. The second project is, in effect, a study 

 of a type of engine which may have just such desired characteristics. 

 When two dissimilar conducting materials are joined to form a 

 loop and the two junctions are kept at different temperatures, heat 

 flows into the loop at the hot junction, a portion of its energy is 

 converted to electrical energy and the rest flows out of the cold 

 junction as heat. The phenomenon involved here has itself long 

 been known; many investigators have been led to speculate upon it 

 as a possibility for large-scale thermoelectric power production, but 

 then to dismiss it as unimportant because the effect is small. Of the 

 early experiments in this field, the best yielded an over-all efficiency 

 of conversion of energy from gas to electricity of only 0.6 percent. 

 Consequently, until recently, thi>, sole use of the phenomenon has been 

 in the measurement of temperature. 



In trying to better these results, one naturally asks, first, the ques- 

 tion "What property must a metal or alloy have besides high thermo- 

 electric power if it is to be of interest for heat-power generation?" 



