THE SELECTION OF GRASSES AND CLOVERS 27 



never yet produced twitch, although it is quite possible that 

 the seeds of that pest may have been sown with an impure 

 sample. 



The misuse of Cocksfoot may be cited as another example. 

 Instead of being reserved for those soils and purposes for which 

 it possesses an undoubted value, it has been recommended for 

 the production of high-class hay everywhere, and even for 

 sowing on geological formations for which it is totally unsuited. 

 Excellent pastures have been literally ruined by the introduc- 

 tion of Cocksfoot. Alternate leys on the Chiltern Hills, where 

 Rye Grass and Clover had previously answered well, have, by 

 the introduction of Cocksfoot, yielded almost unsaleable hay, 

 and, having once been allowed to seed, the Cocksfoot has 

 proved difficult to eradicate. Those who are familiar with the 

 art of forestry are well aware that it ^vould be futile to 

 attempt to gi'ow elm timber on sandy land, or larch on land 

 which is only fit for Scotch fii-. And if an opinion were pro- 

 mulgated that any single kind of timber should be gi*own upon 

 every estate in the kingdom, it would excite general ridicule. 

 Had the laying do\vn of grass received the careful attention 

 which has been devoted to forestry, it would be considered just 

 as unreasonable to sow any one species of grass on all soils and 

 for all purposes, as to recommend elm trees to be planted 

 everywhere. 



The sowing of Poas has been condemned as unnecessary 

 because it happened that some varieties of Poa were indigenous 

 where the experiment was made. But a New England farmer 

 will not hesitate to sow Poa pratensis alone, and long experi- 

 ence has proved that he does not prize this grass too highly. 

 On the other hand, in certam districts of New Zealand, where 

 Poa pratensis develops a troublesome twitch-like habit of the 

 worst character, the suggestion that it should be sown at all 

 would only provoke an expression of incredulous amazement. 

 Surely such a widely different estimate of the value of a single 

 variety may well suggest a doubt as to the universal adaptation 



