253 



spread considerably further than the plants produced from com- 

 mercial seed. 



In 1906 an excellent opportunity presented itself at Cockle 

 Park for testing wild white clover. A small area of the poorest 

 type of boulder clay soil in Tower Hill field had been summer 

 fallowed in 1905, and was sown with wheat in the autumn. In 

 April, 1906, four one-quarter acre plots were marked off and two 

 of them sown with the following seeds (per acre): 



Plot I Plot 2 



Perennial rye-grass . . . 6 Ib. 6 Ib. 



Italian rye-grass 6 6 



Cocksfoot 6 6 



Timothy 3 3 



Meadow fescue 8 8 



Red clover 4 4 



Alsike clover 2 2 



White clover 4 4 



Wild white clover .... 4 



Plots i and 2 had exactly the same seeds' mixtures, except that 

 Plot 2 had, in addition, 4 Ib. wild white clover seed. The seeds 

 for Plot i cost about 235. an acre, and as the wild white clover 

 seed for Plot 2 cost is. 6d. a pound, the seeds for Plot 2 cost 295. 

 an acre. These plots have now produced hay for three years with 

 the following results (per acre): 



"Weight of hay. 

 Plot i Plot 2 



1907 . . . 30 x / a cwt. 35 cwt. 



1908 . . . 18 Y 4 28 V 4 cwt. 



1909 . . . 15 x / a 21 3 / 4 



Average . 21 I / 2 28 1 / 2 



The aftermath has been grazed every year. White clover, and 

 practically all the clovers, disappeared from Plot i after the first 

 year, but now some natural clover plants are spreading on this plot. 

 Plot 2 has always had a thick and close sward of white clover, 

 and this continues to be so. It may be noted that, on this cold 

 clay soil, meadow-fescue seed has failed to produce plants. A strik- 

 ing result is that in Plot i the grasses have not been nearly so 



