(d) It is noted that the essentials in the calculation of machine rates are all based on estimates which may or may not prove correct; that positive book figures are lacking; and that the calculations are made on assumptions. One of the best professional opinions noted is that expressed by one of our leading accountants, who, commenting on the question of machine rates, observes that "it begins with estimating and is estimating all the way through." This appears to be rather severe criticism, yet one has but to give the subject careful consideration to note that it quite correctly sums up the situation in a few words. As previously stated, the machine─rate method is without doubt theoretically correct, but, until the subject has been more thoroughly elucidated and worked out in all its details to fit shop conditions and furnish a satisfactory means of distributing expense, it is doubtful if it will be used to any great extent. This method certainly requires an elaboration of system and detail, with questionable results, and with many serious objections apparent; it seems pertinent, therefore, to ask the question "Is it worth while; cannot something better be devised?" =10. The Percent Method.= A third plan of distribution, commonly referred to as the _percentage method_, differs from the two already outlined, in which the time employed was the basis of operation, in that the rate of distribution is a percentage on the direct labor cost of the product, which is, of course, commensurate with the amount of time expended. In the first division of our subject, it was shown that expense figured on direct labor would prove the most reliable.