Another important change was in the nature of a reversal of the original arrangement, the cylinder or mandrel carrying the record being mounted in fixed bearings, and the recording or reproducing device being fed lengthwise, like the cutting-tool of a lathe, as the blank or record was rotated. It was early recognized that a single needle for forming the record and the reproduction therefrom was an undesirable arrangement, since the formation of the record required a very sharp cutting-tool, while satisfactory and repeated reproduction suggested the use of a stylus which would result in the minimum wear. After many experiments and the production of a number of types of machines, the present recorders and reproducers were evolved, the former consisting of a very small cylindrical gouging tool having a diameter of about forty thousandths of an inch, and the latter a ball or button-shaped stylus with a diameter of about thirty-five thousandths of an inch. By using an incisor of this sort, the record is formed of a series of connected gouges with rounded sides, varying in depth and width, and with which the reproducer automatically engages and maintains its engagement. Another difficulty encountered in the commercial development of the phonograph was the adjustment of the recording stylus so as to enter the wax-like surface to a very slight depth, and of the reproducer so as to engage exactly the record when formed. The earlier types of machines were provided with separate screws for effecting these adjustments; but considerable skill was required to obtain good results, and great difficulty was experienced in meeting the variations in the wax-like cylinders, due to the warping under atmospheric changes. Consequently, with the early types of commercial phonographs, it was first necessary to shave off the blank accurately before a record was formed thereon, in order that an absolutely true surface might be presented. To overcome these troubles, the very ingenious suggestion was then made and adopted, of connecting the recording and reproducing styluses to their respective diaphragms through the instrumentality of a compensating weight, which acted practically as a fixed support under the very rapid sound vibrations, but which yielded readily to distortions or variations in the wax-like cylinders. By reason of this improvement, it became possible to do away with all adjustments, the mass of the compensating weight causing the recorder to engage the blank automatically to the required depth, and to maintain the reproducing stylus always with the desired pressure on the record when formed. These automatic adjustments were maintained even though the blank or record might be so much out of true as an eighth of an inch, equal to more than two hundred times the maximum depth of the record groove.