Symons's Meteorological Magazine, December, 1906, pages 201-205:

By  W.  A.  S.  DOUGLAS,  M.A.
Fig. 1 THE first Wireless Telegraphy Station to be erected in Scotland was completed early in 1906 at Machrihanish, a small village, well known to golfers, on the shores of the Atlantic, some eight miles north of the Mull of Kintyre and five miles west of Campbeltown. This tower was blown down on the afternoon of the fifth of December, during a heavy gale from the north-west, and I send a description based on my familiarity with the locality and on reports in local newspapers. The authorities in charge have given no information on the subject.
    In order to understand the nature of the catastrophe it is necessary first to consider the tower and its mode of construction. The station was built by the National Electric Signalling Company of Washington, U.S.A., in order to exploit a discovery of Professor Fessenden regarding wireless telegraphy. By his method, Professor Fessenden claims that messages can be sent right across the Atlantic, and the tower was designed for the purpose of communicating with an American station near Boston. In order to insure perfect similarity in the two stations, all the machinery and materials for the station in this country were shipped from America and put together at Machrihanish. The buildings, as shown in the accompanying sketch (fig. 1), consist of a power house, with large engines and dynamos, a receiving and transmitting room, and, lastly, the lofty stalk or tower which has now been destroyed by the wind.
    The height of the tower was 420 feet, and the whole structure was built of hollow steel tubes of the uniform diameter of three feet. Each tube was eight feet in length, but as there was three feet of overlap for rivetting purposes as many as eighty-four tubes were used in the construction of the tower. An iron ladder ran up the inside to the top and at each hundred feet there was an outside balcony. The foundation was the most novel feature in the whole construction. The base of the erection, as shown in fig. 3, was a large square of concrete, on the top of which insulators were placed, and then another layer of concrete, in which was embedded a pivot, and upon this the tower itself rested. Therefore, the tower was free to move in any direction, and when it collapsed no injury was done to the foundation. The stalk was held in an upright position solely by means of stays made of stout wire cable. Four stays were attached to the tower at each of the four balconies and the ground ends were fastened to solid concrete foundations (fig. 2), situated at distances of from one to three hundred yards from the base of the tower at points to the east, south, west and north-west of it. By these means it seems to have been thought that the resistance of the tower to the wind would be greatly diminished and any danger of its buckling obviated.
    I have myself been up the tower during a stiff breeze, and even then the swaying was considerable, and the noise of the wind, as it rushed past, was most alarming and indicated that a great resistance was being opposed to its passage. Fig. 2
    For several days before December 5th the wind on the west coast of Scotland was strong and gusty, and Machrihanish always has its full share if any wind is about, but the tower stood the strain well and no one was anxious about its safety. Early on the morning of Wednesday, December 5th, the wind increased greatly and soon was blowing a severe gale from the north-west. As the day advanced the gale became worse and there were frequent squalls of most exceptional severity. About noon the storm seemed to reach its height, and the tower, which was necessarily exposed to the full force of the gale, was observed to be swaying considerably. Everything, however, held fast, although the roar of the wind against the stalk and stays was almost deafening. Shortly before one o'clock in the afternoon the gusts became even more severe, and, during one of exceptional velocity, without any warning the stays to the west of the tower were wrenched from their fastenings and immediately the huge erection collapsed bodily, snapping in two as it fell. The noise of its snapping, together with the crash made when it struck the ground, was heard a long distance away, and at once informed the whole village of what had happened.
    Fortunately, in falling the tower cleared the buildings, though only by a few inches, and some workmen who were in the power house had a very narrow escape. It is worthy of notice that the stays were arranged so that if any one of them broke, the tower would fall clear of the buildings, and this is what actually happened. As already stated, the stalk buckled in its fall, and the two parts fell in different directions. The first section of one hundred feet, measuring from the base, fell against the wind in a northerly direction, while the top section, composed of the remaining three hundred and twenty feet, fell towards the south. Part of the tower fell on a mound, which caused it to snap yet again. The upper portion is flattened out, owing to the momentum of its fall, and parts of it are buried more than a foot deep in the ground. Many of the tubes, needless to say, are damaged beyond all hope of repair, but it is believed that it will be possible to use some fifty per cent. of them again. Fig. 3
    It is characteristic of the energy of those responsible for the station that the manager has got instructions from America to proceed at once with the reconstruction of the tower. This may be taken as an indication that the company's officials are still hopeful of realising the object for which the station was established. It is understood that the experiments, although encouraging, did not come up to the expectations of the inventor, but everything is being conducted so quietly that it is difficult to get trustworthy information. During the summer months, from one till four in the morning, messages were being sent off daily from the station, but how many reached the other side of the ocean it is impossible to learn. The effect at Machrihanish was most remarkable, and it was well worth rising early to see it. Underneath the six acres of ground, leased by the company, a network of wire has been laid, miles of wire having been used for this purpose, and the ground has been railed off with an ordinary wire fence. When messages are being sent off the whole place seems alive with electric sparks: sparks run along the wires of the fence, sparks seem to spring out of the ground, sparks fly through the air from all directions, and their crackling noise mingled with a dull subdued roar can be heard a quarter of a mile away.
    It is unfortunately very difficult to say how the accident happened. So closely are the secrets of the company kept, that it is impossible to find out what pressure of wind the tower was built to withstand, and of course at Machrihanish such things as anemometers are unknown. We can, therefore, only approach the subject indirectly by considering three questions.
    First, was the velocity of the wind greater than could have been reasonably anticipated? A north-westerly gale at Machrihanish, coming as it does over an uninterrupted stretch of ocean, is not a thing to be despised, and I was always inclined to agree with the local opinion, that the tower would never survive a winter's gales, but this opinion was entirely unscientific and may be worth little. It is surely not too much to assume that the tower was put up to withstand a greater gale than ever has blown, or is ever likely to blow, on our shores, and we may therefore answer this question in the negative. *
    Second, was it owing to a mistake in calculating the pressure which the wind would exert on the tower? Without doubt, the tower was put up to resist an enormous pressure, but it seems to me that the effect of an unsteady pressure, such as would be exerted by violent gusts, could not have been sufficiently considered, and that this was the primary cause of the accident.
    Third, was defective material used in the construction of the tower? Possibly this should be answered in the affirmative, for the accident undoubtedly happened owing to some of the stays giving way, and this certainly suggests that a defect in one, or some, of the stays may have been the immediate cause of the whole trouble.
    * The wind on December 5 was undoubtedly very strong along the whole of our western seaboard. At Malin Head, 60 miles due west of Machrihanish, the Meteorological Office observer estimated it at force 10 of Beaufort's scale at 8 a.m. It is a familiar joke that force 12, the maximum of the scale, can only be recorded when observer and observatory have been blown away.
--Ed. S.M.M.