1902.] 



Influence of Carton Dioxide on Leaves, etc. 



407 



mitted to air containing 11*4 parts per 10,000 of C0 2j inflorescence was 

 almost totally inhibited. With the exception of one or two sickly looking 

 flowers on the Begonias, not a single flower-bud opened on any of the 

 plants of this set. The plants of Pmpatiens, Kalanchoe, and of the 

 darker-leaved Fuchsias, did not even produce a flower-bud, whilst in 

 the Nicotiana, CucurUtas, and lighter-leaved Fuchsias, the small flower- 

 buds which commenced to form were completely shed long before the 

 time of opening. 



The plants which appeared to be most sensitive to the action of the 

 «xtra C0 2 in the earlier stages of growth were those of Impatiens platype- 

 ■tala, which from the commencement presented a strong contrast to the 

 healthy appearance of the controls. They lost nearly all their leaves, 

 and the bare stems shed many of their internodes, which were cut off 

 successively just aboA^e the nodes. By June 29, however, the plants 

 had begun to recover, and, by putting out a second growth of small 

 .and very dark green leaves, indicated a certain limited adaptation to 

 their abnormal atmospheric environment. 



So far only the more striking visible differences between the two 

 .sets of plants have been described. It is highly probable that these 

 will be found correlated with equally well-marked differences of ana- 

 tomical structure. This part of the enquiry has been kindly under- 

 taken by Professor J. B. Farmer and Mr. S. E. Chandler, who will 

 embody their results in a separate communication. 



In another series of experiments, which we carried out on similar 

 lines, the air of compartment B was enriched with carbon dioxide to 

 the extent of 6 per cent., that is to say, up to about 200 times the 

 normal amount. The plants used in this experiment were as fol- 

 lows : — 



Nicotiana sylvestris. 



,, tomentosa. 

 Solanum giganteum. 

 Fuchsia. 



Picinus communis, green variety. 



Pdcinus communis, red variety. 

 Chrysanthe mum Broussonettii. 

 Begonia multiflora. 

 Kalanchoe rosea. 

 Cucurbita Pepo. 



The experiment extended from June 3 to August 26, and the 

 general results both in the direction and amount of change of habit in- 

 duced in the plants were so very similar to those induced in the plants 

 with only three and a-half times the normal amount of C0 2 , as to require 

 no further special description. The results are, however, valuable as 

 indicating that the observed differences cannot be due to any direct 

 poisonous influence of the carbon dioxide, otherwise we should certainly 

 expect a marked difference to be produced by increasing the amount 

 of C0 2 from 11*4 parts per 10,000 to 600 parts per 10,000, i.e., more 

 than fiftyfold, which was not the case to any appreciable extent. 



The direction in which we must search for the true explanation of 



