From The Strange World of the Crossword by Roger Millington
If The Times crossword puzzle is the world's most famous, then by common consent the world's toughest are those that have appeared weekly for almost fifty years in The Observer. Only The Listener can claim to rival this series for consistent ingenuity. This supremacy was won for The Observer first by the great Torquemada (Edward Powys Mathers) and then by his successors.
Mathers was born at Forest Hill in 1892, the only son of Edward Peter Mathers, the founder of the newspaper South Africa, and of Mary Powys, through whom he was related to the three Powys brothers, T. F. Llewelyn and John Cooper. In 1924, while earning his living as a literary critic, he came across the crossword puzzle which had then just reached England. At this time crosswords on both sides of the Atlantic employed only the simplest 'dictionary clues' but Mathers realised that with more difficult clues it held the makings of a first-class entertainment. While he may not have invented the 'cryptic clue', there is no doubt that he was the first compiler to use cryptic clues entirely. For the amusement of his friends, he constructed the first Torquemada puzzle, writing the clues in couplets, in the style then current in The Observer. Against his wishes, a friend took it to the Editors of the Saturday Westminster, who persuaded him to produce more. In all, Torquemada published twelve puzzles in this paper; their appearance was heralded by green posters bearing the warning 'Crosswords for Supermen'. Later the twelve were collected in a book entitled 'Crosswords for Riper Years'.
When the Saturday Westminster came to an end, Torquemada was approached by The Observer to contribute puzzles on similar lines. The first was printed in March, 1926, and appeared under the title 'Feelers' as Mathers felt he was feeling his way with his new and wider public. Over the next few months he received a massive correspondence, much of it from readers protesting mildly that they were wasting many hours tussling with his clues. Within a couple of years he already had his copyists employing cryptic clues; at the same time his own 'constant solvers' were getting on his nerves and requesting a fuller display of his invention and ingenuity. He then abandoned the normal crossword grid pattern and devised a form without any black squares. This gave him greater elasticity in the choice of words and enabled him to reduce to a minimum the number of unchecked letters. Only his successor Ximenes had such mastery of the art of composing tortuous and exasperating - but always scrupulously fair - clues. During the years that he worked on The Observer puzzles, he received many letters from solvers indulging in speculations as to his identity. His fondness for Biblical clues led many to endow him with ecclesiastical rank. In fact, Edward Powys Mathers won considerable eminence under his own name as a poet and translator. Even the verse in which he composed his puzzles was often quite distinguished: one of his solvers wrote to say that she had learned many of his rhyme puzzles off by heart.
Enquiries were also frequent as to how he set about composing his weekly puzzles. Several years after his death, his widow answered this curiosity in the foreword to a book of Torquemada puzzles: 'I see him sitting cross-legged in bed, with a puzzle in front of him, looking very like a somewhat relaxed Buddha, a cigarette between his fingers and eyes fixed in the distance - until something clicks and, with a contented smile or discontented shrug, he writes on the list in front of him, and ticks off the word in gaily coloured chalk. Or prowling around his shelves in baggy flannel trousers, his shirt open at the neck and sleeves rolled above the elbow, in search of a quotation through which he would lead his solvers to read or reread some favourite in verse or prose. Or sitting at a table in the living room, kitchen or garden, one ankle resting on the other knee, a hand hugging the foot, drawing marginal decorations in vari-coloured chalks
while he broods on some uninspiring word.'
How long did it take him to compile his masterpieces? According to his widow, the more straightforward puzzles took on average about two hours - although he rarely completed one at a sitting, preferring instead to divide into quarters and sand-wich it with other work. Puzzles with the clues buried in a narrative story or those based on a particular book or author took longer because of' the amount of preparation required. Although his remarkable fertility led many solvers to believe that a Torquemada team was at work, his only collaborator was his wife. Mathers would choose his subject and make a list of words he wished to include. H is wife's part was to select from this list and construct the diagram. From time to time, readers would post him their own 'Revenge' puzzles and occasionally he would borrow a clue from one of these, but less than fifty of the thousands of clues he presented came from these contributions. When his widow examined some thirty thousand clues in The Observer series, she found the same word cropping up fifty times over the years and was astonished at how he succeeded in continually varying the clues.
Considering the difficulty of his puzzles, the wonder is that so many readers were able to solve them. Up to seven thousand correct solutions were received by The Observer each week, and it was estimated that another twenty thousand regularly com-pleted the puzzle without bothering to put the result into the post in pursuit of the prizes for the first three correct solutions opened each week. However, on occasions there were only a handful of successful solvers - on at least two occasions the list of prize-winners was restricted to one lady. Torquemada addicts were widespread. Solutions were received from a man in West Africa who didn't even have a dictionary to turn to. The first air-mail post from India brought a solution, while another came from four men snowed up in Alaska with only a copy of The Observer for entertainment. A Scottish lady of over seventy relied on com-pleting them before Morning Prayer, otherwise her worship was distracted. On the other hand, on one occasion the entire Balliol Common Room admitted that working in combine they still hadn't managed to finish one particular1y brain-twisting puzzle.
A great many solvers worked together in concert sometimes over the telephone: an anguished complaint came from a Scot bewailing the expenditure on trunk calls over one set of clues.
Up to his death in February, 1939, Mathers published 670 Torquemada puzzles in The Observer. It is a great pity that these have so long been out of print for so many years. A selection of 112 Torquemada puzzles was printed in book form in 1942 - Torquemada: 112 Best Crossword Puzzles, published by the Pushkin Press. More unusual is The Torquemada Puzzle Book, published by Victor Gollancz Limited in 1934. This contains a number of crossword puzzles with a Cheats' Dictionary con-taining a list of all the words used in the puzzles, but without definitions. The Torquemada Puzzle Book also includes a section of perforated tracing papers; the idea is that you can tear these out, lay them over the puzzle diagrams and work out your solutions. This way, the unmarked book can be enjoyed by more than one reader. The book finishes with a 100-page detective story called 'Cain's Jawbone'. What makes the story so special is that the pages are printed in the wrong order. Each page has been written so as to finish at the end of a sentence and readers were invited to work out the correct page sequence. Despite the offer of a cash prize, only three correct solutions were received by the publishers!