1890.] On the Question of the Fixation of Free Nitrogen. 95 



with a little distilled water. The roots were then spread out upon 

 paper, and so photographed, and finally noted upon. 



Enlarged photographs of the roots of the plants grown in pot 1, 

 with the yellow sand without soil-extract seeding, in pot 2, with the 

 same sand and soil-extract seeding, and in pot 4, in the garden soil, 

 were exhibited. 



The roots in pot 1, with the yellow sand without soil-extract seed- 

 ing, showed a densely matted mass of fibre, those of the different 

 plants being considerably interwoven ; and although a few fibres 

 reached the bottom of the pot, and distributed through the flints, by 

 far the greater portion was accumulated within the top 4 inches of 

 the sand ; and, notwithstanding there was here no soil-extract seeding, 

 there were many nodules on the roots, but they were fewer, and 

 generally much smaller, than on the roots grown with soil-extract 

 seeding ; they were also less characteristically accumulated near the 

 surface, and more distributed along the root-fibres. There were, 

 however, some agglomerations of nodules. Comparing this result 

 with that obtained in 1889, with a purer and sterilised sand, 

 there can be little doubt that the development of nodules, and the 

 comparatively luxuriant growth, in this pot without soil-extract 

 seeding, are to be attributed to the impurity, and non-sterilisation, of 

 the sand. 



The roots in pot 2, with soil-extract seeding, also showed a dense 

 mass of fibre, which, however, extended from the top to the bottom 

 of the soil, penetrated the layer of flints, and distributed over the 

 bottom of the pot. In fact, the roots were much more generally 

 distributed throughout the soil, and less accumulated within the 

 surface layers, than in pot 1. The most developed root of the three, 

 had three large agglomerated nodules, each with some scores of pro- 

 tuberances, somewhat as on a raspberry or mulberry. The other 

 plants also showed similar nodules, but of a smaller size. There 

 were also a number of small clusters distributed over the rootlets, but 

 very few single nodules, differing in this respect from the character 

 of development observed in pot 1. 



In pot 3, also with soil-extract seeding, each of the three plants had 

 developed a mass of root-fibre extending throughout the soil from 

 the top to the bottom ; though the greatest quantity was within the 

 first 6 of the 7-J- inches of depth. There were large agglomerations 

 of nodules on the roots of each plant. There were, besides, many 

 small clusters, and here and there single nodules. By far the most 

 of the nodules were within the top 3 inches of the sand ; but one 

 considerable bunch was found as low as 4 inches from the surface. 

 As in the other cases, the nodules were grey, and much lighter in 

 colour than the roots on which they grew. 



Each of the three plants in pot 4, with the garden soil, had a 



