forced down with a pressure of about five tons to the square inch; the pressure being continued until the whole had cooled down, and the powder had formed into a solid mass.

The cylinder was then placed over a hollow block of iron as shown at E, in Fig . 8, when pressure was again applied to the plunger and the cylindrical mass of plumbago was forced out, after which it was found to be in every way suitable for making the very best lead pencils.

A young friend of mine to whom I showed some of this compressed plumbago, offered to purchase the invention, at the same time saying that he could not risk more than £200 on the venture; I, remembering the rebuff with the sawing-machine, accepted his offer without further consideration; my friend then went off to Cumberland, and made arrangements with the Plumbago Company. At the present day we find that the best lead pencils in the market are made by crushing the small lumps and odd pieces of plumbago, then washing and floating the powder, by that means getting entirely rid of the little black diamonds, and producing various grades of hardness by different degrees of heat and pressure.

I fear this little episode does not speak very favourably for my business capacity in those early days, for I certainly ought to have made much more than I did by this really important invention.

When I was experimenting with plumbago (about 1838) I was engaged in designing a new system of casting types by machinery, some features of which are of sufficient interest to be recorded. The moulds in this machine were entirely composed of hardened and tempered steel, shaped by laps, as the metal could be neither planed nor filed. From fifty-five to sixty types were cast per minute in each of the two compartments of the mould; and in order that the solidification of the metal should take place in the extremely small interval of time allowed for that purpose, the moulds were cooled by a constant flow of cold water through suitable passages made in them, in close proximity to those parts where the fluid metal came in contact. Another special feature of this mode of casting was the employment of a force pump placed within the bath of melted metal, by means of which the latter was injected into the mould at the proper moment, the pressure of the injected fluid being under the perfect control of a loaded valve. It will be readily understood that a sharp jet of fluid metal would propel with it an induced current of air, and consequently produce a bubbly and spongy casting, which would have been wholly valueless. The short space of time occupied in its solidification afforded no opportunity for the escape of air in the usual way by floating in bubbles upward, as in the case of castings where the metal is retained in its molten state in the mould for several minutes.

I found an absolute cure for this apparently insuperable difficulty, by forming a vacuum in the mould at the very instant at which the injection of metal took place; and so successful was this system of exhausting the moulds, that one might break a hundred types in succession without finding a single blowhole in any one of them.

The iron or brass founder, whose slow and tedious operations are performed by quietly pouring his molten metal into the mould with a ladle, will at once see what a new departure in the art of founding this machine presented. Firstly, there was the same mould producing fifty-five to sixty castings per minute, instead of being broken up and destroyed after one cast: then pouring the metal from a ladle was replaced by injecting it with a force-pump, the mould itself having a continuous stream of cold water running through suitable passages formed in it so as to cool every part of its surface in contact with the fluid metal; and, finally, instead of the mould being composed of porous materials through which the confined air gradually escaped, there was an almost indestructible mould, wholly free from pores, from which all the contained air was withdrawn in the fraction of a second by its sudden connection with an exhausted vessel at the moment when the metal was injected.

The valve through which the metal was injected into the mould being extremely small, required to be fitted very closely to prevent its leaking; it was found that after it had been opened and closed some six or seven thousand times, a portion of the fluid metal would, by friction against the sides of the valve, be rubbed into powder, and more or less obstruct its action. Otherwise, the really beautiful mechanism