318 EXPERIMENTS ON TUENIPS. 



went into a covered dung pit, where it was turned and mixed, 

 and kept entirely under cover till carted out to the drills. The 

 manure for the experimental plots was taken from a homo- 

 geneous section across this dung pit. It is evident from these 

 experiments that nitrogen can be applied to the turnip crop so 

 that it will return a much greater quantity in increase for 100 

 lbs. applied than when given as a heavy dressing of dung. At 

 the same time, as was to be expected from the growing period 

 of turnips continuing later and their roots having a longer 

 possession of the soil, more of the nitrogen from dung is taken 

 advantage of by the turnip crop than by a cereal crop. In the 

 case of barley, the Eothamstead experiments have shown that 

 only 10*7 per cent, of the nitrogen in dung applied is recovered in 

 increase. The relatively better effect of dung upon turnips may 

 also be partly due to their being more dependent upon a supply 

 of carbon in the soil than cereals. In wet seasons, such as 1882, 

 the mechanical effect of the dung is also very beneficial to the 

 turnip plant on this clayey land in keeping the soil open. As, 

 however, a large quantity of the nitrogen in dung — at least three 

 fourths, whether calculations be based on the composition or 

 cost of the dung — is not exhausted by the turnip crop, and 

 remains subject to loss in bare ground for several months, it 

 would probably be more economical to apply the dung to grass 

 whose roots have possession of the land all the year, and would 

 reduce the loss of nitrates in drainage water to a minimum. 

 The average results for some years of these experiments on 

 turnips must be supplemented (1) by observations on the effect 

 of the unexhausted manure on future crops, and (2) by experi- 

 ments upon the effect for a succession of years of one application 

 of dung to grass land in the quantity and quality of the pro- 

 duce, before a definite statement can be made as to the relative 

 economy of these two methods of applying dung. 



IV. Influence of Manures on Gomioosition of Turni]js. — As 

 before mentioned, the turnips were not analysed in 1882. The 

 turnips selected for analysis in 1883 were intended to be the 

 average size of the bulbs on the plot. After thorough cleaning 

 they were weighed in the laboratory, and were found to 

 represent relatively the weight of the crops as nearly as could 

 be expected. It was evident, however, from their weight and 

 the distance between the bulbs, that all the samples were 

 slightly above the average size. On this account the absolute 

 amount of dry matter per acre on all the plots is more likely to 

 be slightly under than over stated, there being generally more 

 water in large than small bulbs. Slices representing the bulbs 

 were dried till they could be finely powdered, and the powder 

 weighed and preserved in air-tight bottles. The residual water 

 was determined by passing dried air over the sample kept at 



