202 



Jfrnnrriiif/ for ffaif 



year to year and also at different periods of one and the same season'. 

 This would account for the fact that Cameron, taking samples in May 

 and June, found much lower proportions of the late flowering grasses, 

 such as bent, than were found in the later analyses based upon samples 

 taken in July. Apart from differences of this character the various 

 observations show su])stantinl agreement. The results of the later 

 analyses are suniinariscil in Table III. 



Table III. 



Name 



1. Lolium peremic 



2. Alopecurus pralensis . . . 



3. Daclylis glomerala 



4. Ar'ena flare-scens 



5. Poa trivialis 



6. Poa prntensts 



7. Fc.'iliini nriiin et dririuscula 



8. Arrhcmillicrnm avetmceum 



9. Anihuxanihum odoralum. 



10. Agrostis vulgaris 



1 1. Brnmiis nwllis 



12. Holni.i luiiatus 



K{ Rinnt'T nrrtosa 



14. Hcrath inn sjfjiondijlium . 



15. Anl/irisrii.s .si/lirslrifi .... 

 10. Bunium flexuosum 



1 7. lAizula campeslris 



18. Various other weeds ... 

 111. Lalhynts pratensis 



20. Trifoliuvi prateiise 



2 1 . Undetermined 



Number of Plot 



1 2 



% 

 •I 



2-8 

 3-5 



8-2 



•2 



1-9 



M 



3-3 



53-4 



2-4 

 ir)-4 



2-2 

 •2 

 ■03 



4-8 



/o 



1-2 



Ifil 



14-9 



4-3 



3-6 



•1 

 31 



•2 

 13-6 



•6 

 1 0-1 

 10-3 

 10-2 



•1 



•05 



5fi 



% 



1-5 



5-3 



120 



34-3 



1-9 



•1 



1-3 



3-8 



•2 



1-4 



1-9 



3-2 



24-3 



■5 



3-5 



■4 



■01 



•2 



•1 



41 



4* 5* 6 



% 

 10 

 61 

 5^8 



204 



■1 



■1 



19 



14 



•9 



37 



14-3 

 0-2 



34-4 

 11 



■1 

 10 

 10 



■5 



% 



20 



1^6 



108 



242 



•5 



1-6 



•2 



19 



328 



7 



211 



1^0 



1^2 



■1 



•04 



0/ 



/o 



■7 



90 



19^9 



204 



20 



■1 



■9 



9-8 



■02 



M 



2^7 



■8 



18-6 



102 



■3 



•4 



■1 



30 



0/ 



/o 



•9 



20 



297 



31-9 



•1 



■2 



4^6 



■2 



■2 



12^6 



■3 



2-4 



91 



■9 

 ■03 



38 



8* 9 I 10* 11 12 13* 



% 

 12 



■1 

 4^8 

 72 



•6 



•4 

 30 



■2 

 142 

 58^8 



■2 

 50 

 14 



■3 



•2 



•7 



■4 



12 



% 



1-2 



38 



268 



134 



•2 



•1 

 2^6 

 84 



■8 

 90 



•7 



22 



207 



21 



I % 

 11 



59 



15 



■5 



•4 



[5^ 



8^5 



71^4 



■1 



21 



2-3 



1-8 

 ■01 

 ■01 



62 



■2 

 ■2 

 •1 



■03 



■G 



/o 



18 



11 



27-8 



70 



■2 



7-4 



■1 



11 



234 



■4 



17 



147 



20 



21 

 ■3 

 •1 



■01 



/o 



14 



•4 



123 



•7 



% 



•6 



•1 



33 



1^6 



■5 — 



11 1-4 



■5 — 



45 57 



573 740 



•II 1 



11 le 



10-2 4^1 



•5 - 



111 54 



•51 •? 

 •l! -8 



■05 



7 7 6 



* Analy.sed in 1911 by Ruston. All other plots analysed in 1909 by Bond. 



It will be noted that the dominant grass on tiie unmanurcd land 

 (IMots 1 and 13) is Agroslis vulgaris, but that this has been very largely 

 suppressed on the dunged plots (2-6). It persists to a considerable 

 extent on the plots manured with artificials only (7-12), being much 

 less pronounced on the nitrate plots (7, 9, 11) than on the ammonia 

 plots (8, 10, 12). Rumex acetosa is ]ii(iiiiiii(Mil throughout and is 

 increased rather than diiiiini.shed on tlic duiigcil plots. It is least 



' Cf. Armstrong, This Journal, 1907. ii. 2'.t'.l. 



