422 



NATURE 



[December 25, 1919 



Important as is the position of the fine old pastures 

 of England in the agricultural economy of the 

 country, and interesting though it may be to examine 

 questions of seeding, a much more important line of 

 inquiry is opened up by the problem of the improve- 

 ment of our second- and third-rate pastures, \^^hat 

 proportion of the grass-land of the country falls into 

 the lower categories it is impossible to say, but the 

 most superficial acquaintance with rural England is 

 sufficient to carrv conviction that the a'ggregate area 

 of such land is enormous. Most of the poor grass- 

 land of the country is associated with the heavier 

 classes of soil, and' has been abandoned to grass on 

 account of the high costs of cultivation, including, m 

 manv cases, the necessity of drainage. It is, for 

 arabie purposes, essentially wheat-land, with an occa- 

 sional crop of beans, and the regular intervention at 

 comparatively short intervals of a bare fallow. Other 

 areas of poor pasture, smaller in aggregate extent 

 than the clays, but still of much importance, are to 

 be found on all the geological formations of the 

 countrv. Of the 14,500,000 acres of permanent grass 

 in England and Wales, 70 per cent, is under pasture 

 and onlv 30 per cent, under hay, and of the poorer 

 classes of grass-land it is certain that the proportion 

 that is grazed is still greater. It is evident, therefore, 

 that the improvement of pasture is relatively a more 

 urgent matter than the improvement of meadows, 

 though with more than 4,250,000 acres of permanent 

 rtrass made into hav in England and Wales during 

 T918, the latter problem is also one of enormous 

 importance. The most famous experiments on the 

 effects of manure on permanent hay are those started 

 in 1856 bv Lawes and Gilbert on the meadow at 

 Rothamsted, and continued ever since on the lines 

 originally laid down. The results have thrown a flood 

 of li«ht on the principles of manuring, which has 

 been 'of the greatest assistance in the elucidation of 

 problems in agricultural chemistry and soil physics. 

 Thev have also shown unmistakably the effects of the 

 more important elements of plant-food on the yield of 

 hav and on its botanical composition, but, even sup- 

 ported as they were bv elaborate chemical analysis 

 of the produce, they leave us uncertain in regard to 

 the feeding value of the herbage. 



\ verv large number of experiments have been 

 carried o'ut which had for their object the determina- 

 tion of the quantitative results attributable to the use 

 of manures, singly and in combination. In many 

 cases these experiments were supported by a botanical, 

 and not infrequently bv a chemical, analysis of the 

 resultant herbage, but it was felt that we were still 

 in a state of much uncertainty in respect of the 

 Quality of the hav— that is to say. its effect on animals 

 consuming it. This induced Middleton in the winter 

 of looo-i to carrv out a feeding experiment with 

 ■sheep at Cockle Park, and in 1005-6 and 1007-8 Gil- 

 Christ continued and amplified this work. The sheep 

 were accommodated in a special house. The various 

 lots of sheep all got equal quantities of roots cake, 

 and hay The hav employed \vas the produce of 

 variousl'v manured olots on old grass-land which I 

 laid out' in 1897- The soil is a clay loam on a boulder 

 clav subsoil. This set of experiments includes the 

 ei«ht-plot test, and it mav be interesting to see what 

 influence nitrogen, phosphoric acid, and potash 

 "es^ec^Wely have on the produce. The quantitatiye 

 figures refer to the average annual yield for twenty - 

 one years. 1897-1017. while the figures which indicate 

 the relative values of the produce, as determined bv 

 the live-weight increase of sheep, are based upon the 

 feeding test'J already specified. The hav from the 

 unmanured plot. No. 6. is assumed _ to be worth 

 ""per ton. The results are set out in the accom- 

 nanvins table ; — 



NO. 2617, VOL. 104] 



Value per ton 



Averaae annual ol hay as 



Plot Manuring per acre per annum yield of hay determined 



by feeding 

 cwt. J. d. 



6 Unmanured 19^ ^ ° 



7 30 lb. >i in Sulphate of Ammonia 23 72 o 

 » 50 lb. P0O5 usually in Basic Slajj 20 tjl o 

 9 501b. K2O in Muriate of Potash... 16 So o 



10 3olb. N+solb. FjOs 3oi ^4° 



n 30 lb. N + 50 ib. KjU 21 720 



12 5olb. PjOs + solb. K2O 26 1019 



13 3olb. N-t-solb. PjOj-t-solb. KjU 30^ 892 



Nitrogen derived from sulphate of ammonia, and 

 used at the rate of 30 lb. per acre per annum, has 

 consistently increased the yield and as consistently 

 reduced the quality. When used alone the nitrogen 

 has increased the crop by 3I cwt. per acre, and 

 reduced the feeding value of the hay by Ss. per 

 ton. When added to phosphates, the nitrogen has 

 increased the yield by 45 cwt. and reduced 

 the quality by 95. per ton. When nitrogen was 

 added to potash the yield has been raised by 

 5 cwt. per acre, and the value lowered by 85. per 

 ton. When used as an addition to both phosphates 

 and potash the nitrogen has increased the yield by 

 44 cwt. per acre, while the value has fallen by I2X. 7d. 

 per ton. Even if the quality of the hay be dis- 

 regarded, the use of nitrogen has always been 

 attended bv an adverse financial balance ; when quality 

 is taken into account this undesirable result is greatly 

 emphasised. 



As regards phosphoric acid, an increased yield has 

 been consistently obtained by its use, accompanied in 

 every case bv a marked improvement in the quality 

 of the hav. 'Taking the arithmetical mean, the in- 

 crease in 'quantity has been nearly 8^ cwt. per acre, 

 while the increase in quality is represented by i6s: per 

 ton. 



The behaviour of potash is rather peculiar. It has 

 quite distinctly reduced the yield when used alone or 

 when used in' combination with nitrogen only, while 

 in both these sets of circumstances it has had no 

 influence one wav or other on the quality of the hay. 

 When added to phosphates it has proved powerless 

 to increase the yield, but it has raised the feeding 

 value of the hav'bv 8s. Q,d. per ton. When added to 

 both nitrogen and phosphates the potash has been 

 practically inooerative so far as yield is concerned, 

 but it has improved the quality by 55, 2d. per ton. 



These results show that verv erroneous conclusions 

 mav be reached if, in experimental work on meadow 

 hav, attention is given only to the weights of pro- 

 duce secured. Thus, in these Cockle Park experi- 

 ments, on the average of twenty-one years, if quantity 

 alone be regarded, sulphate of ammonia used by itselt 

 has involved an annual loss of 6,?. \i. per acre, 

 whereas, if the reduced Quality of the hay be taken 

 into account, the loss is increased to 15.';. -^a. per 

 acre On the other hand, a quantitative gam ot 

 4S. 2d. per acre per annum from the use of phosphate 

 and potash is raised to one of -^is. 5d. owing to the 

 sunerior quality of the hay. While there ,s a certain 

 relationship between the chemical composition, the 

 botanical analysis, and the feeding value of the ha> 

 there will nrobablv be general agreement xv.th 

 Middleton when he "says that "without an ^PP^f ' !° 

 the animal, the relative values of samples gro^^n 

 under different treatment cannot be measured. In 

 mv view, this form of research may, with advantage, 

 be largely extended. 



Turning now to the improvement of pastures as 

 contrasted with meadows, it may be remark-d thar 

 while no sharp line can be drawn between th«e t^" 

 classes of grass-land in respect of ameliorative treat- 



