222 Mr. R. H. M. Bosanquet on the 



practical purposes. For (1) the error can only affect each 

 resulting division once, i. e. it does not appear in the entire 

 turns : and (l } ) the effect of the error on a given division is 

 divided by 180. So that, if as be the maximum error of the 



micrometer-wheel, — ^-r is the maximum error of the resulting 



division. In the above case this furnishes a division abun- 

 dantly accurate for ordinary purposes. 



Arrangement for Turning large Disks and Circles. — A pair of 

 small bearers is provided, standing in the middle of the lathe- 

 board at right angles to the principal bearers. The lathe-head 

 can be lifted off its place and put down on these small bearers. 

 The mandrel is then at right angles to the principal bearers, 

 and distant from them 2 or 3 inches. In this position disks 

 or circles up to 4 feet in diameter can be attached to the man- 

 drel and turned, the slide-rest travelling on the principal 

 bearers along the face of such disks or circles. By aid of this 

 arrangement I turned and divided the circles of my preces- 

 sional globe, which I have since exhibited to the Royal Astro- 

 nomical Society. 



The lathe was said to be ready for delivery in January of 

 the present year. On account of the masons' work then going 

 on, I was unable to receive it at that time. Ultimately, when 

 it was delivered, it turned out that some of the fittings had 

 been forgotten, and others needed revision; so that the lathe 

 was not entirely fit for use till well on in the spring. The 

 acoustic arrangements not being as yet in a state of forward- 

 ness, I thought it legitimate to devote some time to the pro- 

 duction of, first, the small model precessional globe, and after- 

 wards the larger precessional globe, which I required for the 

 investigations in ancient astronomy with which I have been 

 concerned. This, with the preparation of chucks and tools of 

 various kinds for the lathe, and the construction and adjust- 

 ment of the electromagnetic dial hereinafter described, occu- 

 pied me from the time that the lathe was ready up to June in 

 the present year. 



Among the tools constructed at this time is one which I 

 think it worth while to describe more particularly, as I have 

 never seen any thing like it, and its uses are manifold. I 

 could not have divided the globe, or turned any of the larger 

 work done on the lathe, without it. 



It is a solid bar of iron, 1 inch xx 3 in cm 



square, expanded at one end into a $^x> ) — j 



tool-holder, the length of which is 

 at an angle of 30° with the length 

 of the bar. There are screw-holes 



