A good description of how this was done in is found in a 1905 article in Geyer's Stationer (vol. 39, Feb 16, pp. 7-10) which profiles Waterman's hard rubber factory:
After being thoroughly rolled the material is ready to be worked into shape. In order to make a barrel or cap the rubber, thoroughly warmed so as to have it pliable, is rolled on what is called a mandrel -- a shaft or spindle of iron with centers in each end upon which hollow work such as this rubber is driven for the purpose of turning the exterior. To form the closed end of the cap or barrel a small piece of rubber the size of the mandrel is placed at one end; then this little cap and the mandrel are wrapped together several times with a thin sheet of the rubber. Being hot, these parts of the rubber are thoroughly cemented as soon as they come together, leaving no air spaces and no leakage. This leaves the cap and the barrel perfectly tight at the lower end. The covered mandrel is then placed in the vulcanizer and subjected to steam heat of the proper temperature for a number of hours. After leaving the vulcanizer the metal rods are withdrawn from the rubber, and it is then ready for the turning lathe. The original mandrel leaves the hole the proper size for either barrel or cap, but the outside is made a trifle larger than the finished holder. This allows for turning down and removing any blemishes in order to give the surface a high finish.The rough finish often seen on the inside of slip-on caps of the era would be consistent with manufacture as described, with little to no interior finishing after vulcanization. The account is somewhat inconsistent in that it first implies that the mandrels used in vulcanizing are also used to hold the parts for exterior turning, but then describes how the mandrels are withdrawn before sending the parts to the turning lathe. The latter procedure does seem more likely in a production environment, however, and especially if the machining operations took place in a location removed from the molding and vulcanizing work. Removing the mandrels would be much easier with the rubber still warm, and it would make more sense to leave another mandrel mounted on each lathe used for finish turning, onto which the blanks could be press-fitted.
It may not be so easy to see how the closed ends were formed when examining pen parts of unfaded black hard rubber. It is much more obvious with patterned, colored, or faded hard rubber, where breaks in the patterning and subtle color differences show where the end disk was inserted. In the big red hard rubber Williamson cap above, there is even a small gap where the material didn't join completely prior to vulcanization. While some pen components were made by cementing (or even vulcanizing) an end plug into pre-made tube stock after vulcanization, in such instances the plug will be perfectly round and precisely centered -- not the case here.
Certain pen parts were nearly always made by being molded to shape prior to vulcanization. Taper caps, for example, are easily made by molding, but are tricky to machine out of solid stock. Oversized caps and barrels likewise can be molded with great efficiency, while hogging them out of solid rod is both time-consuming and hugely wasteful.
UPDATE: There is a much more detailed (and better illustrated) article on Waterman's manufacturing methods in the December 1911 issue of Machinery, pp. 249-53. And there is a companion article on the manufacture of gold pens (nibs) in the January 1912 issue, pp. 377-79.
