On a recent visit from the family one granddaughter remarked astutely that I had rather a lot of books. There was no arguing with that. Then came the question that I should probably have anticipated: “Which one is your favourite, Grandad?”

Kalaglow: a victim of one of the notable errors in the General Stud Book when his dam and another mare had their identities mixed up

“Well, see if you can find it,” I replied. “I’ll give you a clue. It’s in my study, and it’s about five yards wide.” She looked puzzled, but went to look all the same, then came back, insisting not only that I had no such book, but that she could not believe any book would be that big.

I had to show her, pointing to everything on three wide shelves, and was promptly accused of cheating. So I explained that some books come in more than one volume, and that the one which was my favourite came in 48 volumes, some of those in more than one part.

I could tell she wasn’t convinced, but after I had shown her that the title page on the first volume was the same as on the last volume I detected some grudging measure of acceptance. Of course, she then she had to ask: “What is the General Stud Book, anyway?” I felt that could wait for another day.

Still, I had given an honest answer to her first question. The GSB has been my favourite book since I acquired the newly issued Volume 35 in 1965 and decided that somehow I must contrive to get my hands on all of the previous 34, then keep my set complete by purchasing future volumes.

The development of blood-typing allowed more confidence in the authenticity of GSM entries, and once DNA testing had been introduced the possibility of errors in identification was pretty much erased

What makes the GSB so special for me? Well, I love history and Flat racing, and more particularly the history of Flat racing through breeding, so I’m bound to cherish the work that encapsulates it all. It charts the development of the thoroughbred virtually from its origins in the late 17th century – Old Bald Peg and all that – and I never cease to marvel at the inventiveness, the dogged determination and the diligence of the fellow who conceived the plan and established the model that would be emulated by his successors wherever the sport flourished all over the globe.

That man was William Sidney Towers, who set himself the Herculean task of trying to record all that breeders had done over the best part of a century, and to end the evil of false pedigrees, which had been commonly bandied about. His first stab at the job was published in 1791 and the outcome was remarkable. He had to depend on receiving reliable information from those he canvassed and from records previously published, but there was obviously much that was simply unavailable to him. So much had gone unrecorded.

Strange though it may seem, a significant amount of information that he could not gather is becoming available now. A band of amateur sleuths constantly search documents and archives relating to the nobility and gentry who were once prominent in breeding that have been preserved on the web and elsewhere.

Towers included nothing in his publication that he felt could not be authenticated. Today’s enthusiastic researchers are not discovering much that he got wrong; they are mostly filling in the gaps that have existed all along because 18th century breeders failed to keep proper records or omitted to pass on the information they had.

For me Towers has always been something of an unsung hero in the history of the thoroughbred. It should not be forgotten that he was gathering all his data half a century before the law required the registration of human births, marriages and deaths, his pioneering efforts ensuring that the pedigrees of today’s thoroughbreds can be traced back further than those of all but a tiny proportion of the genus homo sapiens.

But we have long been aware that the pages of the GSB contain errors, some of them in quite recent volumes, uncovered and corrected during my time as a keen student of thoroughbred pedigrees. One of the most notable concerned Kalaglow, a top-class performer trained by Guy Harwood in the early 1980s. He was a four-year-old before it came to light that his dam and another mare had their identities inadvertently switched ten years earlier. The matter was cleared up in time for him to run under the correct pedigree in the 1982 Eclipse and King George, both of which he won.

Further back in time there were notorious instances of deliberate deception, none more so than the case of the horse who finished first in the 1844 Derby masquerading as Running Rein. He not only ran under the wrong pedigree, but was a year older than he was supposed to be. The ruse was fortunately exposed and the perpetrators did not profit from their criminal activity. Other names like Francasal and Flockton Grey spring to mind as subjects of fraudulent switches of identity.

Scientific advances, while ensuring the integrity of present-day pedigrees, have been able to detect errors in the identification of horses from the past

The development of blood-typing allowed more confidence in the authenticity of GSB entries, and once DNA testing had been introduced the possibility of errors in identification was pretty much erased. Today’s thoroughbreds compete under pedigrees that are guaranteed correct.

But scientific advances, while ensuring the integrity of present-day pedigrees, have been able to detect errors in the identification of horses from the past. The method devised by Towers and adopted by all other official stud books focused on the female of the species and her produce – the logical way to extend the record from one generation to the next.

When the Australian Bruce Lowe came along a century later he traced pedigrees back to the earliest recorded female source and assigned numbers to the families. He drew a lot of wrong conclusions from his researches, but others accepted his numbering system and expanded on it, dividing the families into branches. The notion that such a classification is actually useful in helping breeders to produce better horses is largely discredited these days, but once geneticists understood that the transmission of mitochondrial DNA was a female-line function, they had the ability to check the authenticity of the numbered families.

We soon learnt that there were anomalies, and that the faults would not lie in the MtDNA. In numerous cases differently numbered families were found to go back to a common ancestress, and a few years ago there came proof that the horse who won the 1880 Derby as Bend Or and became sire of the mighty Ormonde was not Bend Or at all. He was Tadcaster, as was asserted by some at the time of his stellar career on the turf.

The work on the x (female) chromosome that uncovered so much was always deemed to be worthwhile, as it was known to contain much more genetic material than the y (male) chromosome. Study of the y seemed to offer little prospect of yielding valuable information, but that view is perhaps due some revision.

A couple of months ago I was contacted by Suzi Prichard Jones, who wanted to draw my attention to the fact that the male line of the Byerley Turk seemed to be dying out in this part of the world. I was aware of its evident decline, but had not realised that in Britain and Ireland it now had no more than five representatives – Dunaden, Indian Haven, Notnowcato, Orientor and Pearl Secret.

What were the chances that one of them might suddenly become commercially attractive to breeders, or that some other representative of the line might be imported and become favoured with good opportunities? Where might such a horse be found?

While we were scratching our heads over that situation, we heard that there had been recent research into the horse’s y chromosome undertaken by a team headed by Barbara Wallner in Vienna. And errors had been found, among them the revelation that male-line descendants of Persimmon and St Frusquin had the markers associated with the Byerley Turk line.

Persimmon and St Frusquin between them won the colts’ Classics of 1896, each a son of the multiple champion sire St Simon, and hitherto recognised as products of the Darley Arabian line. And there is a plausible candidate to explain the anomaly.

St Simon’s sire, the 1875 Derby winner Galopin, was recorded as being by Vedette, a bona fide Darley descendant. But there had always been a suspicion that he was instead a son of Delight, from the Byerley line. We might start to believe that now.