Dr. Gardner's Researches on the Functions of Plants. 427 



rod eudiometer, I proceeded as follows: — The plants when 

 wanted were obtained in a perfect state by immersing the 

 tumbler in a tub of water, and removing the garden-mould by 

 agitating the fluid; they were thus procured without the 

 slightest mutilation. The plant was then transferred to a 

 convenient pneumatic trough, closed from adherent air and 

 broken under a small receiver. This was done uniformly at 

 11 o'clock a.m. and as quickly as possible; the gas was im- 

 mediately analysed. 



8. Binoxide of nitrogen was the only substance which could 

 be used to estimate the oxygen, and if properly prepared, is, 

 in my experience, equal to the most intricate apparatus. Thus, 

 in twenty-five analyses of air, made as test experiments of the 

 excellence of my measures, there was obtained a mean of 

 20*83 per cent, oxygen, from which there never was a varia- 

 tion of 0*2 per cent. ; this result closely coincides with that 

 obtained in the elaborate researches of Dumas and Boussin- 

 gault, i. e. 21*8 per cent. 



9. Six analyses of the internal gas of Datura gave a mean 

 of N 87*5, O 12*5 per cent, without any carbonic acid. 



Four analyses of the gas from grass gave a mean of N 86*1, 

 O 13*9 per cent, without carbonic acid. This result closely 

 approximates to the measure published by Dr. Draper in the 

 Philosophical Magazine for the gas drawn by the air-pump 

 from grass. 



The mean of all the observations is N 86*75, O 13*25 per 

 cent. ; and this I assume as the normal or plant atmosphere 

 of the green parts at 11 to 12 o'clock a.m., during vigorous 

 existence in the presence of bright diffused light in summer. 



10. It is distinctly to be understood that this conditionally 

 normal atmosphere is perpetually changing, and is true only 

 for the time and place. In the preparatory examination of 

 this subject, I obtained measures of the internal gas consisting 

 of N 84*6, O 13*0, C0 2 2*4 per cent., but overlooked the cir- 

 cumstances. Messrs. Calvert and Ferrand {Ann. de Ch. f fyc, 

 Aout 1844) found that carbonic acid was always present at 

 night, and give as the composition of the gas from the hollow 

 stems of Phytolacca decandra at night, N 76*4, 020*6, C0 2 

 0*3 per cent. 



III. The Action of Roots on the Gases of the Soil-Fluid. 



1 1 . There are no observations on the action of roots known 

 to me, except those of DeCandolle (Phys. Veg. t. i. p. 248), 

 who asserts that uninjured roots exhale no gas, either in light 

 or darkness. Most physiologists infer, that whatever gases 

 exist in the soil-fluid are absorbed therewith ; but this is an 



2 G2 



