110 Sir J. B. Lawes and Prof. J. H. Gilbert. [Jan. 9, 



were comparatively few agglomerations or bunches, and more single 

 nodules, pretty widely distributed along the root- fibres. The lupin 

 roots, on the other hand, showed tubercular developments very 

 different from those on either the pea or the vetch roots. Indeed, at 

 the period of examination, that is when the plants were nearly ripe, 

 two apparently distinct kinds were observed ; one of which, the most 

 prevalent, we have spoken of as " swellings," and the other as 

 "nodules." The " swellings" were chiefly on the main roots or the 

 thicker branches ; where they grew they encased the root entirely, 

 and they had a shining and presumably impervious skin. The 

 " nodules," on the other hand, were chiefly single, small, and dis- 

 tributed on the root-fibres. Assuming that the so-called " swellings " 

 were the bodies which, with their contents, had exercised the func- 

 tions of the " nodules " found on the roots of the other plants, it is to 

 be concluded that, after the very luxuriant growth, and the flowering 

 and seeding, their function was so far at an end, and they had 

 become suberised. The other bodies on the lupin roots, distinguished 

 in the description as "nodules," indicated too meagre development to 

 have had much share in the great amount of assimilation that had 

 been accomplished. On the other hand, the " swellings," as has been 

 said, were all on the main roots or thicker branches ; whilst it is 

 generally stated that the nodules are only formed on the young and 

 still growing roots. If these " swellings," which were certainly very 

 characteristic of the roots of the plants which attained the greatest 

 growth, were really the effective nodules, it must be supposed that 

 they had been formed where they were found, whilst the root was 

 still young, and had grown with its growth. In favour of this suppo- 

 sition is the fact that the increased growth from the soil- extract 

 seeding commenced quite early in the life of the plants. 



In 1887, the nodule development on lucerne roots was observed at 

 different periods of the season, and again quite recently, on plants 

 taken from the field for that purpose. The nodules on the roots of 

 lucerne are quite different in general external character from those 

 on any of the other plants that have been examined at Rothamsted. 

 Instead of being more or less rounded, they have more the appear- 

 ance of shoots or buds, much longer than broad, sometimes single, 

 but more often divided, or branched ; there being generally two or 

 three, and sometimes as many as twenty, or even many more, in a 

 hunch, joined at the base. They have not been observed on the main 

 root, but only on the root-fibres, and less near the surface than within 

 the range of £he clay subsoil. In some cases such a tuft or bunch 

 will be at the end of a fine fibre by which it is connected with the 

 main root. As the season advances these bodies become shrivelled, 

 and are in fact empty shells. The question arises, whether in the 

 case of the development in soil or subsoil containing organic nitrogen, 



