39° 



Formation of Permanent Pastures. [oct., 



columns show (i) the weight of seed sown; (2) the percentage 

 of the surface allotted to each species as calculated from 

 Stebler's table ; (3) the composition of the first hay crop on 

 the Boulder clay soil of Saxmundham ; (4) the composition of 

 the hay crop 011 the Oxford clay soil of Abbotsley ; (5) the 

 composition of the hay crop cut at Abbotsley in the fifth 

 season, and after the land had been depastured for three years 

 (it was cut for hay in the first year) ; (6) the percentage of the 

 surface occupied by the different plants in the pasture of 1905, 

 i.e., on grass which had been cut for hay in 190 1 and pastured 

 afterwards. Each of the seed-mixtures occupied one acre, 

 which was divided into four quarter-acre plots. These were 

 (a) unmanured ; (b) treated with 10 cwt. basic slag per acre in the 

 autumn of 1900 ; (c) dressed with 10 tons farmyard manure per 

 acre in 1900 ; while (d) received the residues of cake, fed to sheep 

 in 1 90 1 and 1902. The figures in every case represent the 

 averages of the four plots, except in the last column, where they 

 refer to the land manured with slag. In the case of these three 

 mixtures, the different sub-plots were pretty much alike in 1905, 

 and the figures obtained on the "slag" plots would approxi- 

 mately represent the others. 



A very brief inspection of the table will serve to show how 

 slight the connection between the seed and the crop may be. 

 Take first the clovers. We allot to them 30 per cent, of the 

 surface, and at Saxmundham with a fair May rainfall we find 

 42 per cent, in the hay produced by Mixture II., whereas on a 

 similar soil after a dry month of May we find but 8 per cent. 

 This result is obtained where clovers have been brought into 

 competition with rye-grass. As the competition is removed, we 

 find red clover increasing from 67 to 12*5, and finally to 27 per 

 cent, in the hay of 1901 at Abbotsley. The percentage in the 

 hay is not, of course, the same thing as the percentage of ground 

 occupied, but the relationship is close enough to enable us to 

 see that in the case of this experiment a dry month of May in 

 one instance and the competition of rye-grass in another have 

 determined the quantity of clover in the hay crop. " The plan to 

 scale " has worked out fairly well with red clover in Mixture III., 

 but this is obviously due more to chance than to planning. 

 If we carry our investigations beyond the first hay crop we find 



