Within recent years, Kellner's system of starch equivalents has received much criticism from various quarters, and its value as a practical feeding system has been called into question. That the system is far from perfect in its working is universally admitted, and the recent work of W. Bruce with oxen has done much to emphasise this fact. In a series of practical feeding trials with oxen, Bruce obtained results contrary to those expected from the theoretical consideration of the Kellner system of starch equivalents, and largely on account of this work, and partly on account of the demonstration of the importance of the part played by certain unknown chemical compounds in the normal metabolism of animals, many workers have been led to suggest a total abolition of the Kellner system. It is, however, inconceivable to the author that a system based on 30 years continuous research upon feeding problems, should be condemned without the fullest possible enquiry. Furthermore it is supported by many experimental results. In a critical comparison of the values of certain foods for milk production theoretically computed by the Kellner system, Crowther has shown that the results obtained with one exception agree very well with those obtained in actual feeding practice. Moreover, with regard to the food deficiency factors which play such an important role, in the normal metabolism of animals, Hopkins has shown that those factors are hardly likely to be missing in normal dietaries, and consequently the results obtained in ordinary feeding trials will rarely be affected by lack of these factors.