334 Profs. Percy Frankland and Marshall Ward. 



bottle-screens, filled with saturated alum solution, and a pair of 

 coloured glasses, arranged as follows : 



On one microscope- stage lay an alum-screen over the mirror, and 

 on this transparent screen a blue glass, then the culture chamber. 

 Over the culture came another blue glass, and, lastly, an alum- 

 screen. 



Thus all light which passed into the chamber, whether reflected 

 up from the <mirror or direct from above, must traverse the alum- 

 screen as well as the blue glass. 



By the side of this stood a second microscope loaded in exactly 

 the same way, except that a "dummy" cell with its (blackened 

 bulb) thermometer replaced the true culture cell. 



The third microscope was exactly as the first, but with red 

 glasses in place of blue ; and by its side an exactly similar control 

 with a " dummy " cell. 



Each pair of microscopes was further provided with a suspended 

 thermometer, with bulb blackened, to give the air temperature. 



All four microscopes stood side by side at the south window, and 

 each was supplied with a narrow vertical shutter just broad enough 

 to cast a band of shade across the mirror and loads, and so keep 

 direct rays from the cells. 



The day was brilliantly sunny, with deep-blue sky, and very few 

 clouds till 2 P.M. Then heavy white clouds kept coming up, but the 

 sunny intervals were numerous. 



The following were the thermometric observations, the cultures 

 (mirrors and screens) being illuminated only by light from the sky 

 never direct from the sun in the case of the cultures, and only occa- 

 sionally and for short periods to get experimental information from 

 the "dummy" cells. 



From this table we observe (1) that the culture-cells were never 

 allowed to exceed at most 25 to 26 C indeed the cultures them- 

 selves never went so high, because the sun was not allowed to shine 

 directly on either the mirrors or the top glasses ; and (2) that, on the 

 whole, the temperature of the cell between the red screens was about 

 0'5 to 1*0 higher than that of the cell between the blue screens. 



