Wilson — Food- Unit Method in the Fattening of Oattle. 29 



In order to avoid extremes, only experiments which lasted over 70 and 

 under 170 days have been made use of. 



Certain experiments could not be used because their food- ingredients 

 were too loosely described. " Eoots," for instance, may be of several kinds. 

 In one set of experiments, "roots" were found, from the original paper, to 

 be almost equal quantities of mangels and swedes, and the food-units are 

 calculated on this understanding. In some experiments " common cotton 

 cake" was fed, and, on reference to the compositions given in the original 

 papers, this was found to be undecorticated cotton cake. One important 

 series of experiments could not be used at all, because the amount of straw 

 consumed was not stated. 



One extremely large, probably an impossible, ration, and another set, in 

 which the gains are suspiciously large for the food consumed, have been 

 omitted. 



Modern experimenters usually have the animals fasted before weighing 

 them. It is not certain that the older experimenters had this done, but, as 

 they are likely to have followed the same practice at both ends of the fatten- 

 ing period, the errors resulting from weighing the animals unfasted are not 

 vicious when different sets of experiments are being used together. There 

 are more serious causes of variation which cannot be taken into account, as, 

 e.g., the varying capacity of cattlemen ; but the results in one direction 

 may reasonably be assumed to be cancelled by those in another. This 

 remark applies equally, no doubt, to the experimenters themselves, and to 

 some extent reduces the value of the deductions to be made from their 

 work. 



When the data had been plotted, as in the diagram, it was obvious that 

 a straight line could not be drawn through the plottings by the eye. Then 

 the average food -units consumed per group and the daily gains produced 

 were calculated and plotted on the diagram in crosses. The spaced line 

 which joins these crosses, however, is neither straight nor regularly curved, 

 and, if it were to be used, the factors which make it crooked would have had 

 to be found and allowed for. Therefore another line must be found. 



The ideal line would be that which should show, for all weights of cattle, 

 the highest possible daily gains and the lowest possible food-units necessary 

 to produce them ; but such a line cannot yet be drawn. The highest daily 

 gain shown in the diagram is o'l lbs., made by a lot of 12-cwt. bullocks, upon ■ 

 19'8 food- units a day; but we do not know that this is the possible limit. 



A Scotch experimenter (Dr. J. W. Paterson) had a lot of bullocks which 

 made 3'66 lbs. a day, over a period of eighty-eight days, upon pasture, cake, and 

 meal ; and, last summer, at Clonakilty Agricultural Station, four bullocks, 



