PERIODICAL LITERATURE 667 



That grass and weeds generally have a dele- 

 Weeds and terious effect on tree growth is axiomatic, but 

 Tree Groiuth the degree of damage has rarely been subjected 

 to experimental ascertainment. Professor Som- 

 erville reports on his own and some others' experiments in this di- 

 rection. The eft'ect of surface conditions, especially of a grass cover, 

 even on old trees which send their roots into the depths is explained 

 by the fact that the trees form a large amount of fibrous roots near 

 the surface, especially humus soils. Different species are more or less 

 affected; generally speaking, conifers less than hardwoods, but such 

 genera as Taxodium, Sequoia, Thuya, and Cupressus less than Pinus. 

 The interesting results of an exhaustive test on fruit trees during 16 

 years on the Duke of Bedford's experimental fruit farm at Woburn 

 are given. The injurious effect of grass was proved not to be due 

 to interference with the circulation of air in the soil or to the amount 

 of carbonic acid gas, or to soil temperature, or to food supply, or to 

 physical condition, or to the microflora of the soil, but to a toxic sub- 

 stance elaborated by the grass. 



Another experiment, by Armstrong and Pratt, in 1912, carried on 

 with forest trees and various weeds, as well as grass, is of interest in 

 showing that the effect was distinct at the end of the first season in 

 the stunted and prematurely shed foliage, shorter shoots than on bare 

 ground ; stinging nettle and creeping buttercup reducing normal 

 growth to one half, grass to one third, or one fourth, or even one 

 eighth, different species of grass having different effect; couch-rye 

 grass and florin proving most pernicious. 



Professor Somerville's experiments confined themselves to 120 

 S-year-old ash trees, planted in 1910, half of them kept free from 

 weeds, the other half sown with a mixture of grass seeds. After two 

 seasons the height of the trees on the first area was nearly 50 per cent 

 greater than that of the grassed trees, in the first year this difference 

 being 31 per cent, in the second year 109 per cent ; and the root sys- 

 tem on the ungrassed area much larger. 



A second experiment with 30 ash trees, planted in ten sets of pots, 

 each sowed with a different seed, excepting two bare check pot series, 

 showed very variable results, which are illustrated in two plates. 

 While the height growth showed but little variation after one season 

 Calthough superior in bare pots), the character of the plants varied 

 considerably, all except the one sown to white clover showing feeble 

 development of foliage ; lucern, gorse, and common broom, however, 



