A little bit of history

A week or two ago on Twitter, Tom Higham posted a photograph of Martin Aitkin and his father, Charles Higham, using Aitkin’s proton magnetometer at Verulamium. As a result, I have been swapping emails with Charles, who now lives in New Zealand and writes a column for the magazine Current World Archaeology. The photograph shows Charles, Martin and Felicity Wild using the machine (Fig. 1).

Fig.1: Charles Higham, Martin Aitkin and Felicity Wild using the magnetometer.

Charles worked at Verulamium with Sheppard Frere in 1958 and 1959. His diary entry for 1st September 1959 reads:

I go on at St Albans until 23rd of this month at least. Today a change. I helped Dr Martin Aitkin… manage his proton magnetometer. This is a machine to measure subsoil disturbance. We tried with partial success to trace a ditch for a 1/3 of a mile.

The ditch is the confusingly named “1955 ditch”, so-called because Frere excavated a section of it under Insula XX, Building 1 in 1955 (see Plate 1b in the first interim report, Antiquaries Journal 1956). This proved to be the early boundary of the Roman town, and on the basis of two sections excavated across it later, was dated by Frere to c.AD 80, and going out of use in c.AD 125. I have mentioned this boundary in many previous posts as it is such a prominent feature of the magnetic survey data.

Aitkin’s survey was also an important one for the development of the method. In 1957 J. C. Belshe had suggested that kilns could be detected magnetically. As a result, Aitkin undertook the first magnetometry survey in March 1958 at Durobrivae (another site we have worked on) and found it was very successful in finding pits and other features. During the next year Aitkin tried the method on 16 other sites. In 1959 he returned to Durobrivae when the A1 was to be widened and would destroy an area of the suburbs. With two machines, each with three operators, they covered four acres in eight hours taking some 900 readings. Having narrowed down the area of interest, they then surveyed a section in greater detail pinpointing kilns, a ditch and other features. The early development of the method, from initial suggestion to a practical tool for archaeology was surprisingly quick.

Aitkin’s discussion of his surveys in the 1959 and 1960 interim reports from Verulamium (Antiquaries Journal for 1960 and 1961) make for interesting reading. In Fig. 2 is the results from our survey for the southern corner of the 1955 ditch in Verulamium Park.

Fig. 2: the southern corner of the 1955 ditch as shown in the magnetic data.

I have marked the line of the ditch with yellow arrows. I was always puzzled as to why published plans of Verulamium showed this corner as a sharp bend rather than the gradual curve seen in our plots. Aitkin gives us the answer:

The disturbance continued in a straight line for 1,100 feet (110 traverses) at which point the municipal cricket pitch had been reached; here, fearful of the groundsman’s eye, traverses were interrupted and continued on the far side of the pitch. However, the disturbance did not reappear there. Measurements made with care on the pitch itself showed that the ditch turned eastwards through an angle of approximately 107°, whence, despite some interference from iron piping, it was followed for 950 feet…

Aitkin, 1960, pp. 22–24.

The area marked marked “disturbance” in Fig. 2 underlies the eastern edge of the cricket pitch and probably represents some hardcore used to level the area. We also found the metal pipe to which Aitkin refers.

Mostly, Aitkin’s surveys were not illustrated, but for the northern corner of the 1955 ditch he produced the following image, here compared with our survey (Fig. 3).

Fig. 3: The northern corner of the 1955 ditch as surveyed in 1960 and 2016.

As can be seen, Aitkin’s survey compares very well with ours undertaken some 56 years later.

Aitkin was very pleased that in 1959 they measured 302 traverses which constitute some 3,000 measurements. Each traverse was usually 50 feet long, with a reading taken at 5-foot intervals. Each traverse took less than two minutes. That is about 5 readings per minute. At that rate I think Martin’s helpers were kept pretty busy! In comparison, when our Foerster is working (which it isn’t at the moment — again!), we collect 3,000 readings every four minutes, with 10cm along the traverses and 50cm between them. This isn’t to boast, but just to show that the developments in the last six decades have often revolved around the speed of collection allowing us to survey greater areas at greater data densities. Aitkin’s work, however, led the way.

Charles is including the story of helping Aitkin at Verulamium in a future column for Current World Archaeology. I look forward to seeing it.