'To the Stability of the Bell Rock Lighthouse'

How the lighthouse was built

Photo by Tony Treger, Arbroath

Saturday, 4th October 1810 - the morning on which the artificers left the Rock for the last time, having completed the building of the Lighthouse.

That evening Stevenson invited his foremen and captains of the Service, the clerk of works, Mr David Logan, and Mr Lachlan Kennedy, engineer’s clerk, and several of their friends, to one of the local inns, where they spent a pleasant evening together, at the end of which a toast was given: “To the Stability of the Bell Rock Light-house” . . .

. . . And no one would be more pleased to see that the lighthouse still stands exactly as they had left it, than the men themselves who had toasted the Bell Rock Lighthouse that evening in Arbroath almost 200 years ago!

Its design and construction can only be described as one of the world's great engineering feats of its time. And yet it is not to Stevenson, or for that matter Rennie, that the whole credit should go for the construction of this great work.

To both these men, the achievement, yes; but it is John Smeaton, the builder of the Eddystone of 1759, who deserves the real credit for the construction technique, for it was he who pioneered the building of stone-built lighthouses in modern times. Both Stevenson and Rennie, however, readily admitted this, and both were continually in praise of this great engineer. The Eddystone was the blueprint for the Bell Rock - but "modernised" with a few new features and improvements which had been developed over the intervening 50 years.

The design of the lighthouse was in itself a masterpiece. This can best be illustrated when one looks at just one area of its construction. The various courses of stones, built layer upon layer and the method by which they were secured, was nothing short of mind-boggling!

Stevenson, however, was confronted with problems which Smeaton didn't have to worry about. The fact that the Eddystone rock was almost always clear of the water even at high tides; whereas the Bell Rock was submerged twice daily to a depth of sometimes 16 feet (4.87 metres), and it was only possible to work on the rock on an average about 2 hours every low tide!

Because of this Stevenson knew that the lower courses of the lighthouse which were covered by high tides could not be left to their own weight and gravity alone to secure them in position. Trenailling and wedging, as Smeaton had done, would be the answer.

Another important feature was the angle or curve with which the base of the tower took off from the horizontal rock surface. The lighthouse, when finally built, had a broader base than some of the earlier designs submitted; the theory being that to minimise the action of the sea against a solid tower in such an exposed situation the force of the sea would be better deflected with a shallower rise from the base than a tower which took off at a steeper angle.

Part of the First Complete Course and the inset showing clearly the holes for the trenails
and the position of the wedges between the blocks

Dovetailing

“. . . by which the stones may be said to hook laterally into one another, forming a vertical bond of connection; but it is on the gravity of the materials that the chief dependence is placed for the stability of the fabric.”

*James Slight (principal mould-maker) and his brother Alexander were responsible for making the moulds for the stones, following the designs made by the Clerk of Works, in this very important aspect of the lighthouse's construction. The moulds were made from slats of wood (fir timber) and strengthened at the joints by strips of iron.

Each stone, whether granite or sandstone, was carefully hewn to this predetermined shape (as per the illustration below); the whole eventually making up a gigantic circular jigsaw of interlocking stones.

All that could be done back in the yard ashore was done! Each course was laid out and carefully checked before shipping out to the Rock. Stevenson sums it up:
"The whole stones of a course had thus to be laid with great nicety, corresponding to a number of checks and marks, previously arranged in the workyard, that the wedges might fit without trouble at the Rock, and preserve the respective positions of the superincumbent courses, and make band throughout the whole fabric."

Gauge ruler or Trainer

The implement used to position accurately the stones, mainly the lower or solid courses. Stevenson, in his own words, describes it:

"A Trainer or Rule, framed of timber, applied by the builders, for ascertaining the exact position of the stones of the respective courses. At one end is an eye or socket on which it was fitted to a steady-pin placed exactly in the centre. This rule was used chiefly for ascertaining the radiating direction of the stones, from the centre towards the circumference, being laid agreeably to corresponding notches and lines marked upon their upper-beds, so as to preserve band throughout the work, and prevent difficulty with the closing or finishing stones".

Stevenson says that on one occasion the Trainer had been left behind in Arbroath by mistake. So much was its importance, they had to send a boat back to the workshop to collect it. It was not possible to lay the courses without it! In modern terms its use could be considered as similar to a type of compass.

Vertical joggles

Each course was secured by a series of cubical joggles (13 in the case of the first complete course) which would act as steadying pins thus guarding again any horizonal movement. These would penetrate the lower course at selected points halfway through the block (ie. 6 inches), as well as into the superincumbent stone of the course above.

Trenails and wedges

The method of trenailing and wedging was carried out exactly as Smeaton had done 50 years before at the Eddystone. Stevenson noted that it was probably the most time-consuming of all the work carried out at the Rock.

Each stone had 2 circular holes (1¾ inches in diameter) for trenails which would eventually pass through the stone being laid, and continue 6 inches into the already-laid stone below. The centre stone had 4 of these holes.

Each trenail measured between 16 and 26 inches in length. At the lower end a saw-cut was inserted into which a small wedge was placed. When driven into position, the lower part of the trenail would split and tighten. It was then cut flush with the top of the superimcumbent stone and another wedge was similarly inserted to the upper end of the trenail and hammered until it split and tightened; thus completing the stability of the stone onto the course below.

Besides the circular holes for trenailling, grooves, oblong in shape, were cut into the sides (or waists) of each block. Pairs of wooden wedges were then driven into the vertical joints of the stones. These wedges were chiefly intended for aligniing up the dovetailed stones before grouting. If Stevenson followed exactly what Smeaton had done, the grooves would have measured 3 inches by 1 inch.

In the case of the wedges, the thicker end of one of the wedge was inserted first, followed the the thin end of the second wedge, and hammered until secure.

Smeaton himself had remarked that by the use of trenails and wedges "no assignable power, less than would by main stress pull these trenails into two, could lift one of these stones from their beds when so fixed, exclusive of their natural weight, as all agitation was prevented by the lateral wedges."

The final bond throughout the entire building was maintained by centering the perpendicular joints of the new course immediately on the middle of the course below.

The above procedure was carried out, not only for the courses below high water level, but also throughout the entire solid part of the building and that section of the between the door entrance (the 26th course) and the floor of the Provision room - in all some 38 courses!

The materials used for the first complete course alone - 123 stones, 13 joggles, 246 oak trenails and 378 pair oak wedges.

The final totals of that part of the house - 1440 stones, 338 joggles, 4065 trenails and 6329 pairs wedges.

"Zonal or belt" construction

From the 39th Course to the top of the House, the trenailing, wedging, and dovetailing were dispensed with (with the exception of the end joints and centre stones of the floors). At this level another method of connecting the courses was adopted, by means of a raised "zone or belt" (as Stevenson called it) fashioned on the upper face of the stone (or more correctly the wall blocks), which would fit into a similar recessed groove in the stone above. This can be seen more clearly in the Courses section.

This and the plan of the horizontal floors, were the chief differences in design between the Eddystone and the Bell Rock.

The lower part of the House showing the change from courses with vertical joggles
to walls-cum-floor/roof using the zonal or belt construction and horizontal joggles

The problems of construction at these levels were perhaps not quite so labour-intensive as those of the lower levels, although the builders at that point had to have more the attributes of steeplejacks than of men used only to working on the ground or by the sea! The great size of some of the stones also worried Stevenson. Any damage or loss to any of the stones (especially those of the floor of the topmost room, (weighing over a ton and about 7½ feet in length) would have caused considerably down-time. Not something Stevenson was willing to contemplate.

Roofs and floors

Another variation from Smeaton, who used chains to secure his arched roofs at Eddystone Lighthouse, was in the construction of the roofs/floors of the courses of the Bell's upper rooms. Stevenson explains:

"The floor courses of the Bell Rock lighthouse lay horizontally upon the walls. They consisted in all of 18 blocks, but only 16 were laid in the first instance, as the centre stone was necessarily left out to allow the shaft of the Balance-crane to pass through the several apartments of the building. In the same manner also, the stone which formed the interior side of the manhole [entrance to the next floor], was not laid till after the centre stone, and the masonry of the walls completed. The number of stones mentioned above are independent of the 16 joggle pieces, with which the principal blocks of the floors were connected [horizontally in this instance, as opposed to vertically in the lower courses].

Note how the curved roof is constructed with huge horizontal blocks to achieve the arched effect
and how the floor of this room also acts as part of the wall

"The floors of the Eddystone lighthouse, on the contrary, were constructed in an arched form, and the haunches of the arches bound with chains to prevent their pressing outwards, to the detriment of the walls. In this, Mr Smeaton followed the construction of the Dome of St Paul's; and this mode might also have been found necessary at the Eddystone, due to the want of stones in one length, to form the outward wall and floor, in the then state of the granite quarries of Cornwall.

"At Mylnefield quarry, however, there was no difficulty in procuring stones of the requisite dimensions; and the Writer [Stevenson] foresaw many advantages that would arise from having the stones of the floors to form part of the outward walls without introducing the system of arching. In particular, the pressure of the floors upon the walls would thus be perpendicular; for, as the stones were prepared in the sides, 'groove-and-feather', after the manner of the common house floor, they would, by this means, form so many girths, binding the exterior together."

Ring of iron

Another innovation for the Bell Rock that would ensure greater stability at that part of the house (the 81st course), considering it had to bear the full weight of the cornice and the projecting stones of the dome, was the introduction of a circular flat-bar of Swedish iron, 3 inches deep by 1 inch wide. For the Eddystone arches, Smeaton had used iron chains instead.

This ring bar (weighing some 400 lb avoirdupois) was fitted into a groove, 4 inches wide by 3 inches deep, which was cut in the upper bed of the course. The bar was then heated to 150 deg Fahrenheit, and sealed into position by molten lead.

To get the entire job done that day, the men worked from 4 in the morning till 8 in the evening - a total of 16 hours.

This technique could be likened to the wright applying the heated outer metal rim to a wooden spoked wheel or the bands of metal on a barrel.

Roman Cement

Also important to the Bell Rock works was "Roman Cement", a type of mortar patented by a Mr Parker of London. Although not considered appropriate for general use, due to its quick-setting qualities, it made excellent grouting for the lower courses and the walls of the entire building where waterproofing was required. This substance (brownish in colour) is produced from calcined nodules or argillaceous limestone, found on the south coast of England. It was also expensive to buy so it was only used when necessary.

The normal mortar mix was made up of lime from Aberthaw, Wales; pozzolano or tarras (an earthy substance akin to lava) from Sicily, Italy, and Andernach, Germany, respectively; also locally obtained sand.

Throughout the actual building of the lighthouse, Stevenson continually sung the praises of his men. When the the last stone was laid, he stated to the gathered artificers and seamen "that towards those connected with this arduous work, he would ever retain the most heartfelt regard in all their interests." Many of the men in Stevenson's Bell Rock team rose in the service. He kept his word!

* George Anthony Slight, who still lives in Chile and is a descendant of the man responsible for making the moulds for the Bell Rock masonry, explained his connection with Scottish lighthouses and the Bell Rock. He says that his ancestor James Slight married a daughter of John Rennie and that his grandfather. George Slight, went out to Chile in 1886 to design and build the Evangelist Lighthouse on the west coast of the Magellan Straits and afterwards stayed on in Chile to build a total of 72 lighthouses.