32 PASTURE STUDIES: SOME RESULTS. 
Agrostis was incemparably more prevalent on Area 7 in later 
years. As Area 7 was grazed with cattle only, at this period, 
it is not probable that a serious amount of the sown seed- 
lings would be uprooted. On the whole, therefore, it does 
not seem probable that the difference was due to this cause. 
(6) The seeds-mixtures for the two areas were quite 
different in somerespects, and the success of Cynosurus 
eristatus especially, and to a smaller extent of Anthoxanthum 
odoratum and Lolium perenne on Area 6 seems to indicate 
that this was the main cause. The development of Dactylis 
glomerata evidently had no appreciable influence against 
Agrostis, as this was even better developed on Area 7 than 
on Area 6. 
Atra caryophyllea. 
This grass is not indigenous to any of the present areas, and 
it was only recorded once. It is known to have been present in 
considerable abundance in the very young pasture of Area 3, to 
which its seed was probably introduced as an impurity in other 
grass seeds.* 
Steglingia decumbens. 
This grass was not recorded on Areas 1-4, but it formed 18-8 
per cent. of the herbage of the old grass-heath of Area 8. It was 
also found in all examinations of Areas 6 and 7, but on the former 
area in relatively very small amounts. 
LEGUMINOSAE, 
In addition to the Percentage Frequency examinations, the 
areas were examined for these plants by the Specific Frequency 
method. 
Trifolium pratense. 
The seeds of a number of so-called varieties of this plant are 
obtainable commercially, and it has been shown that not only 
may these varieties differ considerably from each other in some 
1Its seed is a common impurity of Festuca duriuscula and Cynosurus 
cristatus seed (8), (II), 
