excellently adapled lo Uiis end. Tlie ol)jeclioii usiially raiscd a^ainsl 

 Ihis principle, iiamely that Ihe ohject of ex|)erimenl is exposed lo a 

 continuously lalling (X).^ tensioii, is of lillle imi)orlance,:provide(l llial 

 care is taken that the amoiinl of (X)o ahsorhed is small iii rclalion lo 

 the total amounl of CO.. Paiallel ex|)('riiiienls wilh niy nu-lhod and 

 the aspirator method wilh absorption in a Pi;i TENKori:n luhe have 

 given similar values for the assimilation^ (see e. g. Table I, nos. 12 

 and i;}, Table VIII, nos. 10 and 11, Table IX, nos. O and 7). 



When one has lo work Avilh the exlraordinarily small (pianlilies 

 of CO.^ present in the air, the melhods of analysis used by l'i i:i i-i;n, 

 HoNNiER and Mangix, and olhers are not sufficicnlly exacl. Nor 

 are the gas-analj^sis appaialuses constiucled for dilTerenl purposes 

 by Pettersson (1897), Krogh (19()(S) etc. — though very good in 

 themselves — suitable for eslimations of assimilalion in Ihe small 

 CO., pressure of the air. The apparalus conslrucled by nu*, which 

 combines assimilation-chamber and absorption-room, is arranged as 

 follows. 



The apparatus has the form of a round, open box of /inc. wilh 

 double walls. The space between the walls is lilled Avilh waler by 

 means of the tube (/), the lemperature of the waler is shown by a 

 thermometer. As lid (/) I use a piece of looking-glass. 2".> mm. 

 thick, which rests on a raised edge I cm. broad, smeared wilh 

 vaseline, and is made to seal the vessel hermetically by means of 

 metal clamps with springs (c/). The dimensions are so chosen that 

 the volume of air enclosed measures .'} litres. The ouler walls of 

 the box push up a little above the lid of looking-glass: when Ihis 

 is set on, one can therefore pour waler on to a space 2 cm deep, 

 >vhich is of great advantage for assimilation in sunlighl. Hy means 

 of simple arrangements the waler can be renewed conlinually. In 

 this way over-heating is entirely avoided. The Ihick envelo|)e of 

 water round the box moreover remains, during an experimenl lasling 

 1 hour, more or less at so conslant a lemi)eralure Ihal no s|)ecial 

 heating and regulaling arrangements are necessary, provided that 

 Ihe temperature of the waler is about the same as that of Ihe sur- 

 rounding air. If required, the ouler walls can be covered wilh some 

 heat-isolating material. The raising or lowering of Ibe envelope of 

 water to the desired temperature is elTecled mosl simply by means 

 of a supply of hot or cold water in Ihe lube ,/\ the oullel laking 



* My method liowever gives on au avemge .somewiiat liigher valuo?, a fac-t \Nhi.h 

 is probably dne to the more complete CO,, absorption. 

 4. — SvenKk Botanisk Tiihkrift 19->l. 



