29 



perature was but some tenths of a degree and was over- 

 come very soon afterwards. 



To be able to ventilate without disturbing the thermal 

 equilibrium in the inner vessel, two long spirals of métal 

 tubing were placed in the water-jacket between the 

 inner- and outer-vessels. Thèse spirals (which are not 

 reproduced in ûg. 3) served to ventilate the thermostat. 

 The air used came directly from the roof by means of a 

 pipe and an airpiimp worked by water-power caused a 

 continuons circulation. Some water always remained at 

 •the bottom of the inner vessel so that the latter was 

 saturated with vapour. 



An axle H, with indicator I and graduation K on the 

 side-wall of the outer-vessel, could be rotated from the 

 outside. It turned in two cylindrical cases, soldered water- 

 tight in opposite side walls between the outer- and inner- 

 vessels. On this axle was screwed a copper plate, on which 

 were flve clamps L, so that plants in 5 vessels could be 

 stimulated simultaneously. In view of the experiments 

 on phototropism the inner vessel measured 50 cm. from 

 front-wall to back-wall; the front-walls of the inner- and 

 outer-vessels were of thick plate-glass, and the back-walls 

 could be taken off to détermine the absorption-coefficient 

 of the front-wall. The whole apparatus was made of copper, 

 and whereever the copper came in contact with the water, 

 the surface was coated with tin. 



The theemoregulatob (cf. fig. 4). 



The thermoregulator also was constructed according to 

 the indications of Mr. de G root, mechanician of the van 

 't Hoff laboratory. The regulator consists of an open 

 thermometer with a large mercury-reservoir A and a wide 

 capillary B with a thick wall. The regulator used by me 

 had a bulb of 2 'A cm. diameter and 10 cm. height. Into 

 the réservoir A a platinum-wire had been fused, which 



