SECRET SERVICES are a perpetual subject of public curiosity. Many books have been written about them, mostly ill-informed, many romantic, some pure fantasy. Those who work in them are liable to fantasy too: it is their occupational disease. However, since such services are necessary, we must put up with that, merely hoping that the fantasy will be kept under control by professional skill. Thus controlled, it can even serve as a smoke-screen to protect serious operations.

The British Secret Service takes particular delight in such self-protection. Even its remote origins are hidden behind a distorting fog. However, no fog is quite impenetrable. Most secrets can be discovered if one knows where to look. The central documents may be hidden or destroyed, but there are copies in other files. Private archives preserve private memories and private resentments. The historian needs a hunter's nose and a steady head; but thus equipped, he may find his way even in that dense, dark jungle, so full of strange animals, false trails and distracting noises.

Christopher Andrew is such a historian. Thanks to a rare combination of talents, he has written a history of British Secret Intelligence which is scholarly, balanced, and, at the same time, since he has a lively sense of the absurd (to which, however, he never sacrifices either charity or scholarship), highly entertaining. It is a book which, both as a historian and as a one-time denizen of that jungle, I can praise without qualification.

It is often supposed that the British Secret Service has been continuous for centuries. This is not so. Special services have been created in times of special danger, but then, in periods of apparent security, hve been allowed to lapse -- until complacency was shaken again by some notable failure. In the late 19th century British complacency was shaken by a series of failures first in the Crimean War, then in the Boer War, and by 1900 the sudden discovery that the country was both isolated and vulnerable caused a "great fear." This fear was profitably exploited by a series of "farcical Cassandras" who declared that the country was at the mercy of thousands of spies preparing first for a French, then for a German invasion. Andrew gives a hilarious account of this scare, of its exploitation by the popular writers William Le Queux and E. Philips Oppenheim, and of the amateur, even gentlemanly, response of the establishment. However, it did lead to the creation of the special security service which was afterwards known as MI5. Its creator was Sir Vernon Kell, whom I remember well, for he was still there in 1940. He was finally sacked then -- one of the many victims of Hitler's Blitzkrieg; but he had done good work before 1914, rounding up the score of incompetent German spies who were all that really existed of the 70,000 imagined by the scaremongers to whom he owed his post.

MEANWHILE a secret intelligence service -- the beginning of the modern MI6, alias S.I.S. -- had also been organized. Since the great threat was from the new German Navy, it wore naval uniform. Its first chief was Sir Mansfield Cumming, whose personal idiosyncracies became the regular mysteries and rituals of his office. He was a colorful character, with a touch of fantasy, and lives in office legend; but the great scoops of the First World War were not due to him and his adventurous agents but to the cryptographers directed by Sir Reginald Hall in Room 40 of the Admiralty. These were equally eccentric, but in a different way -- "men of the professor type" as they were condescendingly called -- and their work though never fully exploited, was invaluable: for it provided the "hard" evidence of the German naval order of battle. They also, by deciphering the famous "Zimmermann telegram," helped to bring America into the war. Apart from this, there was strategic intelligence run by the Army high command, whose agents counted trains behind the enemy lines and sent messages by carrier pigeons -- so successfully that the enemy was credited with similar tactics and we are told that it was dangerous for an innocent Englishman to be seen conversing with a pigeon in Trafalgar Square.

Many of the men who ran MI6 in World War I were still there in 1939. Secrecy protects persons as well as information, and in those years the organization, deprived of wartime glamor and funds, acquired some of the characteristics of a privileged club. Luckily the best crypographers were continued too. Transferred from the Admiralty to the Foreign Office, they did valuable work, reading the Russian ciphers: for those were the years of the Comintern and the "great fear" was of Bolshevik subversion. But in 1927, to justify a dubious political gesture, the government (encouraged by the old hands of SIS) published some damaging Russian telegrams. The cryptographers, of course, were furious, for the Russians promptly changed to new ciphers, which were not read. The mind boggles at such a blunder, which, from then on, in a vital area, and in perilous times, left British statesman fumbling in the dark.

In the critical 1930s, SIS was at its weakest. Its eyes, thus willfully blinded, were still fixed on Russia, although the immediate danger now came from Germany. Without "hard" evidence, it had no means of sifting the "soft"evidence that poured into it. Soft evidence can be useful but it needs to be analyzed and analysis depends on intelligence. The comfortable leaders of SIS not only lacked intelligence: they positively distrusted it. Declining, on principle, to recruit university graduates, they left them to be recruited by the Comintern.

The resultwas a series of disasters which continued until 1940. From these disasters SIS was then saved less by its own efforts than, once again, by the cryptographers. Ultra was the great success story of the Second World War. It restored a central pillar to the crumbling edifice of secret intelligence and was vital to victory. Winston Churchill saw at once the value of Ultra and gave his support. SIS saw it too. By hitching itself to the life-line of Ultra, the Old Guard survived the war, basically unchanged -- until the shocks of the 1950s forced, at last, radical reform.

That reform, Andrew thinks, has not gone far enough. In a thoughtful epilogue he argues for greater accountability. I think that anyone who has followed his story will agree. Secrecy is essential to the operations of a secret service; but secrecy, divorced from its purpose and pursued for its own sake, has side-effects which have often been damaging to the service itself.

I am glad that the machinery of such secrecy has not prevented the completion of this well-documented, responsible and very readable book.