184 ACTION OF ROOTS ON GASES. [BOOK n. 



matter, which is so abundant in the spongioles, as immediately 

 to give off ammoniacal vapours when decomposed by aid of 

 heat. Aerial roots, especially those of many species of 

 Pothos, contain more than such as are subterranean. This 

 azotised matter is almost or entirely insoluble in water, and 

 adheres inseparably to the cellular tissue : it is most abundant 

 at the points of the spongioles, and gradually disappears in 

 the interior of the root. It appears essential to the life of 

 plants, and its large proportion at the extremities of the roots 

 may help to explain why azotised manures are so peculiarly 

 efficient. It also shows how the well known destructive 

 effects of tannin upon roots take place, by precipitating the 

 azotised matter, which is essential to the existence of roots. 



Some experiments have been tried by Dr. D. P. Gardner 

 in order to determine the action of roots upon the gases 

 contained in the water which bathes them. 



In June 1844 he commenced his observations upon the 

 uninjured roots of Datura and blue grass. The plants were 

 placed in vessels resembling a bird-fountain, filled with pump 

 water, and capable of being replenished to compensate for 

 the evaporation of the leaves, and also of collecting any gas 

 passing from the roots. " Three sets of experiments were 

 made : A, the roots and leaves were placed in darkness ; B, 

 both portions were exposed to bright diffused light ; C, the 

 leaves were illuminated, but the roots in darkness. On 

 the evening of the 25th of June, two sets of plants were 

 arranged according to these plans. The Daturas B, yielded 

 the next morning at 11, a gas the composition of which was 

 N 96-6, O 3*4 per cent. ; these two plants were then placed 

 in a dark cupboard for thirty-six hours and evolved no gas 

 whatever ; on again exposing them to light, they produced a 

 mixture of N 96-2, O 3'8 per cent, as the mean of six 

 analyses. The grass plants, B, gave off but little gas, and 

 only enough was collected for two measures, which yielded a 

 mean of N 96, O 4 per cent. The plants C conducted them- 

 selves in the same way as B ; the Daturas gave gas for six 

 analyses, the mean of which was N 96*5, O 3'5 per cent. 

 The plants A, placed in darkness, gave no gas whatever, 



