finally: the finials

The Hubbard / Hill Cemetery project required replication of three missing headstone finials for the nearly identical headstones of Hubbard family members who died in 1880, 1881, and 1884. The original finials would of have been made of lathe turned white marble. No evidence of their design was available in photographs or fragments. The Texas Historic Commission warns against making assumptions when replicating missing elements unless there is strong historical evidence to back it up. After some discussion with the owners I felt further investigation might yield some clues and they were game to the idea. The nearby historic Old Red Rock cemetery on highway 821 south of present day Red Rock gave me a good answer. Identical headstones from the same era and of the same dimensions convinced me that the design was a standard monument company model of the period, but only one of the Red Rock headstones had it's finial. Cemetery bandits must love the fact that the tempting tidbit is at eye level and held in place by a single dowel. Unfortunately, over time, the steel or bronze dowel corrodes and becomes loose in it's packing, which may have been either oakum or poured molten sulfer, making the finial as easy to swipe as Jimmy Stewart's loose newel post knob in It's a Wonderful Life. 

The marble finial model was borrowed from it's hole only long enough to measure and photograph it, and then with Jeree's help I traced it's shadow profile (held at right angles to the afternoon sun above a note pad) back at the truck. He was then returned to his proper place - with a little prayer that it doesn't disappear to thieves too.
After having replicas priced by a local stone company (to the tune of nearly 600 dollars!) for the three I thought it was time to come up with our own version. The first step was to replicate the profile in a moldable material. I built a jig - type lathe out of plywood for my reversible drill and turned a replica profile out of dense floral foam to match the dimensions of my notes. The styrofoam was shaped easily to an acceptable facsimile using a small rasp, files, and sandpaper. So far, so good. The foam would disintegrate if sealed with spray polyurethane (discovered from previous experience unfortunately). The texture of the foam would also show up - or not release from a plaster mold - without sealing so I coated the foam with drywall compound, let it dry and turned it again in the lathe jig to a smooth shape with sandpaper. After drying and coating that with several coats of spray poly I made a two piece plaster mold for casting the finials in concrete. A good casting mix for concrete was researched, I tried a 3:2 sand to white portland cement and left it fairly stiff for strength.  The mold turned out fine and left a great reverse shape to receive the concrete but the first casting showed how absorbent the plaster was - sucking the moisture right out of the concrete mix like a dehydrator. The mold was sealed with multiple coats of polyurethane, dried overnight and then coated with a release layer of thinned vaseline. The second mix was made thinner to fill the mold shape easier and left for a good twelve hours to set. The thinnest neck near the base repeatedly broke while unmolding. Different combinations of concrete were tried for moldability, strength, color, and texture. No acceptable results and a table full of broken finals after nearly a week of effort required a different direction. Back to the foam model. Latex molding liquid was applied over several days to 10 layers thick obtaining a flexible rubber mold thick enough to hold up to repeated castings. The profile prevented stretching the mold enough to get a concrete out of the narrow neck so the latex had to be split up the side to peel it off. The latex was stitched back together with masking tape into a complete shape for casting. More concrete - after a day of drying the tape was removed and voila' - a complete finial. BUT, the rubber mold filled with heavy concrete had left the final casting bumpy and lumpy, okay for Claus Oldenberg but not for this project. The latex was taped around the foam model once again and an outer mold was made from plain sacrete just to hold it into shape while the inner castings set up. By this time the mix had evolved to a 1:1:1 mix of sacrete (screened to remove the largest aggregate) to white portland cement to sand and watered to the consistency of thick oatmeal. This allowed a drier mix for strength but still liquid enough to find the inner profile with tamping and tapping the outer mold after pouring. It took 12-14 failed castings to come up with three complete, but far from perfect finials. They were delivered and installed with Barre Pak epoxy onto the patiently waiting headstones. They are intended to represent the originals and they certainly add a tremendous amount of detail to the cemetery. The strong gothic obelisk shape of the headstones look almost art deco in design without them but appropriately Victorian when in place. 

The headstones missing their finial appear almost art deco in design.

From a few feet away it is hard to tell they are imposters. In true preservation form they are not intended to exactly replicate the missing forms but are there to just inform the viewer of the historic detail. Next time I have to do this it should be a cinch ... and now that I have the mold and the process down it is tempting to go on a guerrilla campaign to replace missing finals in other cemeteries; beware all you bastard grave robbers.