740 Plant Breeding at Aberystwyth. [Nov., 



example, that " grassland dairying is not a very perfect prac- 

 tice, owing to the rapid falling off in the milk-producing 

 qualities of the grass after the month of June." For hay, 

 early maturity is a great drawback, especially in districts 

 where the hay is habitually cut too late. Gilchrist has 

 advocated New Zealand cocksfoot on account of its later 

 maturity. The value of meadow foxtail as a meadow grass, 

 also, is much reduced by its particularly early maturity. 

 Second] ji, all the available evidence goes to show that a leafy 

 herbage is more nutritious than is a stemmy herbage ; it is only 

 necessary to cite Hall's and Eussell's work in this connection.* 

 Tlurdhf, it would seem evident that the plan of selecting for 

 maximum output of dry matter per acre as opposed to gross 

 produce per acre would be applicable to grass improvement 

 just as to roots. Thus the aim must be to produce grasses 

 which flower and mature late and which yield the maximum 

 of leafy (as opposed to stemmy) dry matter per acre per 

 annum; and since the temporary ley of 4-6 years' duration is 

 an important feature of grass land management in the West, 

 the ideal grass must also be capable of uniform pt^oductivity 

 over a 4-6 year period. 



Up to the present experiments have been conducted more 

 fully w^ith cocksfoot than with any other grass. Twenty-four 

 different lots, representing indigenous, Danish, American and 

 French stocks were sown in small beds in the spring of 1919. t 

 Each bed consisted of four rows 9 in. apart and 56 in. long. 

 During the past spring and summer hay was cut from the rows 

 on different dates, and the aftermaths cut at regular intervals. 

 It is not necessary here to enter into details of the interesting 

 results obtained, but reference to Table I, w^iich gives the 

 average gross produce from the beds of each nationality, and 

 to Table TT, which gives the gross produce of the best and 

 worst lots, will reveal important facts. 



* See Jour, of A gric. Science, Vol. IV. (4), June, 1912. It is obvious, of 

 course, that "leafiness " is much influenced by habitat, but Hall's and Eussell's 

 data do not preclude the possibility of strains existing with a potentiality for 

 leafiness. 



t It was unfortunately not possible to obtain New Zealand seed for inclu- 

 sion in these trials ; seed from New Zealand has, however, been procured for 

 subsequent sowing. 



