JOURNAL OF THE ROYAL llOUTiCULTUKAL SOCILTY. 



of the experiment it became apparent that the manure was not only 

 affecting the weight of hay obtained, but also the character of the 

 vegetation. Accordingly wdiat we may call botanical analyses were 

 made of the vegetation on each of the plots. When the hay was cut 

 a careful sample was taken that would fairly represent the bulk ; it 

 was then brought down to the laboratory, picked apart plant by plant, 

 and sorted into heaps, each of which represented a separate species. 

 The heap for each species thus obtained was then weighed, so that 



Fig. 1. — Turf from the Plot without Manure for FoRT^r-EiGHT Years. 



Herbage short, showing a great variety of species. Briza media the most 

 prominent grass. 



in the end a percentage table was obtained showing the proportion by 

 weight which each grass or clover or weed contributed to the herbage 

 of the plot. Dr. Masters was invited to co-operate in the study of 

 this field from the botanical side. He spent a good deal of time at 

 Eothamsted in the years 1877-80, and eventually co-operated with 

 Lawes and Gilbert in an extensive paper, which appeared in the 

 Philosophical Transactions " in 1880. 

 The changes which had been effected by the different manures are 

 still more manifest to-day after the treatment has been persisted in for 



