47G MB. J. B. LAWES, DB. GILBBBT, \.\D DB. PTTGB 



As in the case of the soil, o to thai of the atmosphere, the anlj essential conditi 

 to be attained are Buch as are required for health) growth, and as will at the same time 

 enable us to estimate the amounl and the sources of the combined Nitr< • ming 



within the reach of the planl 1>\ it> means. 



In consequence of the mobility of the atmosphere above referred to, it wj 

 to exclude the experimental plants from its free access. The quantity of ammonia in 

 the air is. however, soverj small, that, provided the atmosphere of the enclosing apparatus 

 were allowed to remain unchanged throughout the period of an experiment, the amount 

 of combined Nitrogen so coming within the reach of the plant might be altogether 

 neglected. Nor, so far as regards the nitrogen and oxygen of the air, i- there any 

 necessity for change; but, owing to the peculiar circumstances of temperature and of 

 moisture to which the air of the apparatus is subjected, conditions more closelj allied 

 to those of ordinary vegetation are attained by a frequent change of atmosphere. 'Die- 

 large quantity of air which thus becomes involved in an experiment precluded the idea 

 of neglecting the consideration of the combined Nitrogen which itcontains. It devolved 

 upon us, therefore, either to determine the total amount of combined Nitrogen in the 

 air before and after it came in contact with the plant, or to free the air from combined 

 Nitrogen before- admission into the enclosing apparatus. The latter alternative was 

 adopted as the most simple ; and the manner in which the object was effected will appear 

 from the following description of the apparatus employed. 



H. — TJie Apparatus used to enclose the Plants, and to supply them with Air, W"ter, 



Carbonic Acid, dr. 



Plate XIII. represents the entire apparatus as used for each separate experiment in 

 1S57 ; and fig. 1, Plate XIV., that used, also for each separate experiment, in 1 858, 

 which, as will be seen, several important modifications of the arrangement adopted in 

 1857 were made. 



The same letters of reference apply to the two so far as the parts are alike; and 

 where there has been any modification in the arrangement in 1858, as compared with 

 that in 1857, the same letters represent the parts of the apparatus used for the same 

 purpose in each, with the exception, that those which apply to the modification of the 

 apparatus in 1858, are distinguished by a dash, thus '. 



A, Plate XIII. (and fig. 1, Plate XIV.), represents a large stone-ware "YVoulfe's bottle. 

 18 inches in diameter and 2-1 inches high. 



B, C, and E are glass "YVoulfe's bottles of 30 ounces capacity. 



F is a large glass shade, the dimensions of which were, in most of the experiments, 

 diameter 9 inches, and height 40 inches ; in other cases the dimensions were, diameter 

 16 inches, and height 28 inches. 



a represents the cross section of a leaden pipe 1^ inch in diameter, which is in con- 

 nexion with a reservoir of water, not shown. This pipe passes over all the vessels A (of 



