Out counting deer with the Forestry Commission rangers, I moved into a strange, spooky environment in which heat, rather than light, reveals the secrets of night life on the land. The imager picks up the radiant warmth given off by living creatures, trees and buildings, and is so sensitive that it can detect a rabbit a mile away. Any heat source shows up as a reddish glow.
Cruising in a Land Rover between the woods around New Park Manor - now a hotel, but originally one of the hunting lodges built by Charles I - we soon picked up deer feeding out in a field. The night was pitch-dark, with a thin crescent of moon showing through bare branches, and the naked eye could see nothing. But through the imager even I, a beginner, could discern that the bright pink shapes standing out from the black background were fallow bucks. When we approached within 100 metres, their details became so clear that ghostly flickers of red bounced off their antlers as they turned their heads back and forth.
There were also a dozen rabbits going about their business. As I watched, a fox appeared and made a rush at one of them; when its target escaped, it sat down to take stock of the situation, quite unaware that humans had eyes on it. In this surreal scene, the true colours of the animals were immaterial: because it was their body heat that we were seeing, all showed up the same reddish-pink.
For the young local rangers, Robert Colin-Stokes and Andy Page, this was a trial run, the first time they had used the imager. But they had an expert tutor in the form of Derek Stocker, the commission's wildlife officer for the south and west, who has conducted extensive trials over the past two years.
Counting woodland deer by conventional means is notoriously difficult. Even with a big team of trained spotters, many animals are missed, and totals arrived at are generally reckoned to be one-third below the real ones. In recent years it has become fashionable to count by collecting dung-samples, but this method also has many detractors.
Enter the thermal imager, which started life as a targeting device on surface-to-air missile systems. Two years ago, the Forestry Commission borrowed one from the SAS, and was so impressed that it hired another from the maker, Thorn EMI, for six months. Again, the results were excellent. Now it has bought one of its own, at the trifling cost of pounds 40,000.
Trials in Scotland, Yorkshire, the Marches (in Shropshire) and the West Country have revealed that earlier counts were indeed serious underestimates. For instance, in Mortimer Forest, near Ludlow, the experienced resident ranger, John Speed, reckoned that he had 200 fallow deer on his ground. In a single night with the imager counters he found 460.
The merit of the device is threefold: it is very accurate; it does not disturb the deer, and it enables the counters to sex most of the animals they see. Furnished with precise data of this kind, the Commission is in a far better position to explain to the public why annual culls are necessary (rangers have to shoot about 1,000 deer in the New Forest every year to keep the population stable).
As Mr Stocker found, the view through an imager seems to take some getting used to. The heat patterns given off by sheep and red deer (which have very thick coats) are much the same, so that at long range it is hard to tell the difference. Roe deer, in contrast, stand out very well, and at this time of the year bucks are easily distinguished, because their growing antlers are covered by the thick, hairy skin known as velvet, and this, being full of blood vessels, shows up brightly.
Beginners, however, do not take long to cotton on to the niceties of the new system. "There are 12 fallow does out there," reported Robert at one point during our night cruise. "One of them's just defecated, and the droppings are still glowing red."