Wilh a view lo shoiieniiii^ Ihe lime of aljsorplion as miicli as pos- 

 sible I used j^, solulions. a lilres of normal air (= Ihe ca|)acily of 

 the a|)paiatus) conlain ahout I.7— l.s mg. CO.^, which corresponds 

 to about l.ör. ccm. of J^, oxalic acid. Wilh an accuracy in reading 

 lo O.ou ccm., an eiror of about l.:i % Iherefoie arises. The eslima- 

 lions of air are therefore cerlain at about ^'^ = O.i.i ?6, as far as Ihe 

 error in reading is concerned. In Ihe assimilation however all Ihc 

 amounl of CO.^ should not be used, but only a fraclion of il. In 

 experiments wilh ordinary air I allowed the CO.^ percenlage to be 

 reduced by i to ^^ on the average, which corresponds to about O.r. ccm. 

 of oxalic acid. The error will therefore be 4 %. Wilh raised CO2 

 percenlage and increased assimilation the error is of course relatively 

 less. In a furlher paper I shall describe the resulls of a number 

 of test-experiments wilh known amounts of CO.^. 



Wilh accurale work, therefore, the conslant error should be (|uile 

 inconsiderable in comparison wilh Ihe numerous accidenlal sources 

 of fallacy which are involved in the genelic and modificatory varia- 

 tions of the material, and which must always be taken inlo accounl 

 in ecological work. Il is therefore necessary to make many esti- 

 mations wilh constantly fresh material, and this, thanks to Ihe sim- 

 plicily of the method, it is easy lo do. 



In calculating the resulls one must of course lake into considera- 

 lion the volume of the glass tube gl (fig. 1). 



lUumination. As source of light the natural daylight was used. 

 Some experiments were carried ont in the verv habitals of the plants; 

 but since the light in the forest is subject lo conslant variations, 

 1 carried out the majority of the experiments on the open place 

 outside the Station. 



Close to a wall facing NW, which is in the shade from morning 

 till about 4 ]). m., there could be oblained a slrongly dilVused light, 

 Ihe intensily of which during the months of June, July, and August 

 constituled in clear weather about ]: of the maximum light of heaven. 

 I took as unit the maximum light of heaven, i. e. Ihe intensily in 

 Ihe open lield under full sun at noon in the month of July, 



The light-measurements were carried out wilh an Imperial Ex- 

 ])Osure-meter, No. 1. The experiments during the summer of 1920 

 were favoured wilh good weather. Benealh a sky covered wilh 

 drifting rainclouds the intensily of light varies too much to be used. 

 Lower inlensities than -} were oblained by laying pholographic plates, 

 with varying darkening powers, över the apparaluses. 



