Alan Smith: fitting the continents back together with the help of a computer

Alan Smith tells the story of the origin of and work involved in the first computer aided fit of the continents, undertaken in Cambridge's Department of Geodesy and Geophysics in the early 1960s.

Listen to the full interview track and all other tracks from this interview on British Library Sounds

There was an extraordinary Tasmanian called Sam Carey, some people used to call him the Tasmanian Devil, but he was convinced continental drift had taken place long before it was considered respectable academically in the northern hemisphere. And he wanted to show how well the edges of the continents of South America, and Africa fitted together and you don’t do that at the coastline ‘cause the coastline goes – migrates very rapidly with a small change of sea level; you do it at the edge of the continent where it goes down to the oceanic depths. And he had done this as well as you could possibly do visually, he’d taken – he made a big globe, traced off the bathometric contours, that is the contours at the edge of the continents and going down to the oceanic depths, selected one of them at about – I don’t know whether it was 1,000 fathom or 2,000 fathom.  He’d trace them off very carefully onto plastic caps that he could move around on this globe. He had – he then fitted the two contours of Africa and South America together as well as he could visually and that was his starting point and it really was a very good fit. It looked very convincing to most people and it was shown to Harold Jeffreys who was in this college as the senior fellow at the time and Harold denied that there was a fit, he just said, ‘I deny there is a fit,’ you see. When Bullard heard about this he said, you know, ‘We must do something about this,’ and so what he did was to get hold of a graduate student, Jim Everett and said, ‘Look Jim, you know, can you write a programme that fits these two contours together?’ Which he did, you know, it was a very elegant programme ‘cause the computers were not very large at that time, they were some of the first computers, EDSAC I think was the name of the computer here that did this. And for Jim and – and Bullard and myself when we – the question was not so much 'do they fit together' but 'how good is the fit?' I mean you can different degrees of fit and so we – Jim’s programme quantitatively defined how good that fit was. And I got involved with fitting things together, partly because I had this physical background and understand – and understood the mathematics involved and partly because I'd had some experience with computers at Princeton, because I had been a research assistant in computing when I first got there for one year. And also because I’m a geologist I was the ideal person in a sense to say whether or not you could keep Iceland in the reconstruction, or whether it was plausible to keep Iceland in the reconstruction or whether you could get rid of it. Should you keep Rockall or get rid of it? And I just adopted purely pragmatic point of view with a bit of geology saying, ‘Well obviously Iceland’s in the way, it wasn’t there when they fitted together so we’ll get rid of it, but Rockall looks as if – if we get rid of Rockall we’re going to have a big hole in this reconstruction so we’ll keep it.’ It was almost as basic as that. I mean it was very – almost unscientific but on the other hand essentially what we were trying to do was to fit continents together so that there were no gaps in the fit ‘cause we – I think intuitively we felt at the time that if the contours fitted together they fitted perfectly and they had just been changed – their outlines had been changed as a result of separation. And anything that – having fitted together things as well as possible that lay on top of that spoil the fit as it were we thought there must be a reason, geological reason for that. And in the Iceland it’s obvious now but this is what – we didn’t know that at the time. Erm, so I was able to use my geological knowledge to fit together the North Atlantic whereas Jim Everett had fitted together the South Atlantic and so, you know, we fitted North America to Europe and Greenland and North America to – we had to rotate Spain, close up the Bay of Biscay as Carey had said, and Africa and so on. So we got a – it wasn’t Pangaea quite, it was the major continents around the Atlantic Ocean had been – were fitted together. And the reason I was involved with that was simply because we wanted the best fit possible on which to plot the data we were going to get from dating these rocks from around the margins of Africa and South America. So that’s – that’s basically how I got into this business which is again another substantial part of my research about what the world looked like in the past actually.
  • Interviewee Alan Smith
  • Duration 00:05:05
  • Copyright British Library Board
  • Interviewer Paul Merchant
  • Date of interview 1/5/2012
  • Shelfmark C1379/65

Related Audio Clips

The following clips are short extracts from an in-depth interview.
To listen to the full interview visit

Related Video Clips

Related themes

Related disciplines