44 



out of doors in melting snow, the second one in the 

 thermostat, which easily could hâve caused a slight 

 différence. 



In the case of the experiments out of doors in the 

 melting snow the seedlings were influenced also by the 

 impurities of the laboratory-atmosphere, as the stay out 

 of doors was of short duration only and even during that 

 time the wooden boxes in which the plants were placed 

 out of doors, were filled with the air of the dark room. 



Without any hésitation I therefore accepted the factor 

 'Vg as applicable to ail températures. By ascertaining the 

 fact, that my starting-point (the presentation-time at 20° C. 

 up till May IS"" 4'20", after that date 3') remained con- 

 stant, I could finally bring together ail values as being 

 directly comparable. 



Temperatuee. 



Température plays a very small part as a source of 

 error in this investigation. As has been said above, the 

 thermostat could be kept constant to 0.1° C. In the be- 

 ginning this was done only in cases, in which the plants 

 remained in the thermostat for less than 6 hours; the 

 stirring-apparatus was stopped during the night, when 

 the température remained constant to 0.3° C. Later the 

 stirring-apparatus was also kept running during the night 

 when necessary. At 35° C. — 40° C. ail figures refer to a 

 température regulated to 0.1° C. 



The same thermometer, with a scale divided to 0.1° C, 

 was used in ail experiments. In the course of the inves- 

 tigation this thermometer was compared with a thermo- 

 meter in the Physical Laboratory, graduated to 0.05° C. 



In view of possible errors it was of much importance 

 to know the exact course of the température within the 

 thermostat when heating. 



Tu trace this I madesomethermo-electrical déterminations 



