The laboratory reports stated that the samples taken from all the original eleven horses showed a high adrenalin content, but this was not significant because horses vary enormously, some always producing more adrenalin than others. I, however, think that it was significant that the adrenalin counts of those eleven horses were uniformly higher than average.

The noise which triggered off their fear is the high note of the sort of silent whistle normally used for training dogs. Horses can hear it well, though to human ears it is faint: this fact makes it ideal for the purpose, as a more obtrusive sound (a football rattle, for instance) would soon have been spotted. Humber keeps a dog whistle in the drinks compartment of his Bentley.

I do not yet know for sure how Adams and Humber frighten the horses, but I can make a guess.

For a fortnight I looked after a horse known in the yard as Mickey (registered name, Starlamp) who had been given the treatment. In Mickey's case, it was a disaster. He returned from three days' absence with large raw patches on his fore legs and in a completely unhinged mental state.

The wounds on his legs were explained by the head lad as having been caused by the application of a blister. But there was no blister paste to be seen, and I think they were ordinary burns caused by some sort of naked flame. Horses are more afraid of fire than of anything else, and it seems probable to me that it is expectation of being burnt that Adams and Humber have harnessed to the sound of a dog whistle.

I blew a dog whistle to discover its effect on Mickey. It was less than three weeks after the association had been planted, and he reacted violently and unmistakably. If you care to, you can repeat this trial on Six-Ply; but give him room to bolt in safety.

Adams and Humber chose horses which looked promising throughout then" racing careers but had never won on account of running out of steam or guts at the last fence; and there are of course any number of horses like this. They bought them cheaply one at a time from auction sales or out of selling races, instilled into them a noise-fear association, and quietly sold them again. Often, far from losing on the deal, they made a profit (of. past histories of horses collected by officer cadets).

Having sold a horse with such a built-in accelerator, Adams and Humber then waited for it to run in a selling 'chase at one of five courses:

Sedgefield, Haydock, Ludlow, Kelso, and Stafford. They seem to have been prepared to wait indefinitely for this combination of place and event to occur, and in fact it has only occurred twelve times (eleven winners and Superman) since the first case twenty months ago.

These courses were chosen, I imagine, because their extra long run-in gave the most room for the panic to take effect. The horses were often lying fourth or fifth when landing over the last fence, and needed time to overhaul the leaders. If a horse was left too hopelessly behind, Adams and Humber could just have left the whistle un blown forfeited their stake money, and waited for another day.

Selling 'chases were preferred, I think, because horses are less likely to fall in them, and because of the good possibility of the winners changing hands yet again immediately afterwards.

At first sight it looks as if it would have been safer to have applied this scheme to Flat racing:

but Flat racers do not seem to change hands so often, which would lessen the confusion. Then again Humber has never held a Flat licence, and probably can't get one.

None of the horses has been galvanized twice, the reason probably being that having once discovered they were not burnt after hearing the whistle they would be less likely to expect to be again. Their reaction would no longer be reliable enough to gamble on.

All the eleven horses won at very long odds, varying from 10-1 to 50-1, and Adams and Humber must have spread their bets thinly enough to raise no comment. I do not know how much Adams won on each race, but the least Humber made was seventeen hundred pounds, and the most was four thousand five hundred.

Details of all the processed horses, successful and unsuccessful, are recorded in a blue ledger at present to be found at the back of the third drawer down in the centre one of three green filing cabinets in Humber's stable office.

Basically, as you see, it is a simple plan. All they do is make a horse associate fire with a dog whistle, and then blow a whistle as he lands over the last fence.

No drugs, no mechanical contrivances, no help needed from owner, trainer, or jockey. There was only a slight risk of Adams and Humber being found out, because their connection with the horses was so obscure and distant.

Stapleton, however, suspected them, and I am certain in my own mind that they killed him, although there is no supporting evidence.

They believe now that they are safe and undetected: and they intend, during the next few days, to plant fear in a horse called Kandersteg. I have left Humber's employ and am writing this while keeping a watch on the yard.

I propose to follow the horse box when Kandersteg leaves in it, and discover where and how the heat is applied.

I stopped writing and picked up the binoculars. The lads were bustling about doing evening stables and I enjoyed not being down there among them.

It was too soon, I thought, to expect Humber to start on Kandersteg, however much of a hurry he and Adams were in. They couldn't have known for certain that I would depart before lunch, or even that day, and they were bound to let my dust settle before making a move. On the other hand I couldn't risk missing them. Even the two miles to the telephone in Posset made ringing up Beckett a worrying prospect. It would take no longer for Kandersteg to be loaded up and carted off than for me to locate Beckett in his Club. Mickey-Star- lamp had been both removed and brought back in daylight, and it might be that Humber never moved any horses about by night. But I couldn't be sure. I bit the end of my pen in indecision. Finally, deciding not to telephone, I added a postscript to the report.

I would very much appreciate some help in this watch, because if it continues for several days I could easily miss the horse box through falling asleep. I can be found two miles out of Posset on the Hexham road, at the head of the valley which Humber's stables lie in.

I added the time, the date, and signed my name. Then I folded the report into an envelope, and addressed it to Colonel Beckett.

I raced down to Posset to put the letter in the box outside the post office. Four miles. I was away for just under six minutes. It was lucky, I think, that I met no traffic on either part of the trip. I skidded to a worried halt at the top of the hill, but all appeared normal down in the stables. I wheeled the motor-cycle off the road again, down to where I had been before, and took a long look through the binoculars.

It was beginning to get dark and lights were on in nearly all the boxes, shining out into the yard. The dark looming bulk of Humber's house, which lay nearest to me, shut off from my sight his brick office and all the top end of the yard, but I had a sideways view of the closed doors of the horse boxes, of which the fourth from the left was occupied by Kandersteg.

And there he was, a pale washy chestnut, moving across and catching the light as Bert tossed his straw to make him comfortable for the night. I sighed with relief, and sat down again to watch.

The routine work went on, untroubled, unchanged. I watched Plumber, leaning on his stick, make his slow inspection round the yard, and absent-mindedly rubbed the bruises he had given me that morning. One by one the doors were shut and the lights went out until only a single window glowed yellow, the last window along the right-hand row of boxes, the window of the lads' kitchen. I put down the binoculars, and got to my feet and stretched.