244 Obituary Notices of Fellows deceased. 



and devoted more than two years to the investigation of this subject 

 in the Rothamsted laboratory. The results obtained in experiments 

 with a considerable variety of plants, showed no assimilation of free 

 nitrogen. The chemical statistics of the leguminous crops at Rothamsted,. 

 and elsewhere, showed, however, that they contained an extraordinary 

 large amount of nitrogen, the source of which was difficult to explain. 

 In 1886, Hellriegel and Wilfarth proved that leguminous plants assimi- 

 lated the free nitrogen of the air in considerable quantities, if the soil 

 in which they grew contained certain microbes forming nodules on their 

 roots. The experiments giving rise to this conclusion were repeated at 

 Rothamsted with a similar result. The very different results obtained 

 in the earlier and later experiments at Rothamsted were due to the fact 

 that the plants in the earlier experiments were all grown in burnt soil, 

 and the microbes were thus excluded. 



Among the miscellaneous investigations conducted at Rothamsted, 

 may be named those on the relation between the amount of water tran- 

 spired by plants and their increase in dry matter ; the investigation 

 on the composition of the milling products of wheat grain ; the in- 

 vestigations conducted for the Government, on the manurial value of 

 sewage ; also the chemical study of the " fairy-rings " in meadow land. 



The experiments relating to animals were very numerous in the 

 earlier years of Rothamsted work. Trials were made on a large scale 

 of the comparative fattening capacity of different breeds of sheep.. 

 The sheep were kept under ordinary agricultural conditions, and the 

 relation between food and increase was carefully ascertained. The 

 trials extended over several years. Numerous feeding experiments of a 

 more scientific character were made on fattening pigs these received diets 

 containing very varied proportions of albuminoids and carbohydrates. 

 It was found that the supply of a larger proportion of albuminoids than 

 that contained in cereal grains was not attended by a greater increase 

 in live weight. This conclusion was contrary to the scientific opinion 

 then prevalent, which regarded the amount of albuminoids in a diet 

 as a measure of its nutritive value. Feeding experiments on oxen 

 were conducted by LaAves and Gilbert at Woburn. 



A very important and laborious piece of work was the determina- 

 tion of the percentage composition of the whole bodies of animals, 

 oxen, sheep, and pigs, of various ages, and in various conditions as to 

 fatness. The proportion of all the organs, and of the butcher's 

 carcase, in the live weight, was ascertained in the case of a large 

 number of animals ; and in the case of ten animals, the proportion 

 of water, fat, nitrogenous matter, and ash was determined in every 

 part, and by calculation in the whole animal. The ash was after- 

 wards analysed and its composition determined. The facts thus 

 ascertained still form our chief source of information as to the com- 



