328 



Profs. Percy Frankland and Marshall Ward. 



Tbe coloured glasses used were the ruby ones marked in table. The 

 thermometers in the air were exposed to the sun, and so the tempera- 

 tures given are not really those of the air, but they serve my purpose 

 for this preliminary experiment, which is only intended to give an 

 instance of the enormous variations which may be got if the observer 

 is not on his guard against insolation. 



On looking at these results, one or two interesting facts of consider- 

 able importance to the investigation come out and serve as a basis for 

 great caution in exposures of this kind. 



In the first place it is clear that direct insolation causes such enor- 

 mous rises of temperature in the cells, that no results could be 

 expected sufficiently definite to be of use unless these could be con- 

 trolled. Secondly, it is equally clear that the variations of tempera- 

 ture in such a south window are too sudden and great to be neglected 

 as serious sources of error ; and, thirdly, the ordinary mode of record- 

 ing the temperature by means of thermometers merely hanging near 

 the cultures obviously gives very imperfect results, and cannot be 

 trusted for any such purposes as these. 



On the other hand, the records as given by the blackened-bulb ther- 

 mometer in the dummy and model cells (especially the latter) may 

 be taken as true indications of the temperature in the culture-drops, 

 and it is instructive to observe that the culture-cells respond slowly 

 to alterations in the temperature of the air, bat very rapidly to 



