The material that went into the six episodes broadcast on BBC Radio 4 in 1978 as “The Hitchhikers Guide to the Galaxy”, was already outlined in Adams’ mind no later than February 1977 when he met Simon Brett of the BBC and received a verbal and exploratory commission over lunch. The pilot script was complete by the 4th of April, and recorded in June 1977. On the basis of that pilot a full script was commissioned on the 31st of August 1977, and recording took place in November and December 1977. The first episode was broadcast on the evening of the 8th of March, and the sixth and last on the 12th of April 1978. The series was entirely self-contained and no sequel or reworking in other media seems to have been anticipated by anyone involved. Indeed, Adams had already accepted the demanding job of script editor for the television series Dr Who, and his time was so fully committed that he was even compelled to bring in his close friend John Lloyd to assist with the writing of episodes five and six of The Hitchhikers’ Guide.

However, the success of “Hitchhikers'” was immediate, and without delay the authors, and not just the authors, began to think of new forms in which the existing material could be recast, as well as a sequel. No more than six weeks after the last broadcast, in May 1978, Nick Webb of Pan Books met Adams and Lloyd, who was at this point regarded almost as a co-author, to discuss the possibility of a novella to be published as a mass-market paperback. They quickly moved to terms, and a large advance was paid to Adams and Lloyd before the end of the year. However, after a short period of hesitation, Adams very deliberately moved to exclude Lloyd from the projected book, resulting in a serious though not final breach between them.

In August 1978, the BBC commissioned a Christmas episode, and before long there were plans for a second series, of which parts of the Christmas broadcast were eventually to form the first episode. This follow-up series was expected to go into pre-production in August 1979 and to be broadcast in January 1980. In fact, recording did not take place until January 1980 and the broadcast was delayed until August that year.

In parallel with these developments, but taking precedence over them, Adams was working on the first novella for Pan Books, The Hitchhiker’s Guide to the Galaxy, which was published in October 1979, and then almost immediately on the second novella, The Restaurant at the End of the Universe, which was available in bookshops exactly a year later in October 1980.

For these two books Adams split the material from the six episodes broadcast in 1978 into two sections, but made substantial changes to the plotline at exactly the breakpoint between the two books. The motivation for this revision appears to be largely, though not quite entirely, driven by the wish to remove Lloyd’s contributions, which were substantial in precisely this part of the narrative.

The second Radio 4 radio series is now widely known as the "Secondary Phase", a term that retrospectively and misleadingly baptises the initial six broadcasts as the “Primary Phase”, though as already noted, this was not how they were seen at the time of their recording, when no sequel was contemplated. This “Secondary Phase”, the radio broadcast sequel, is independent and perhaps even subsequent to some of the changes made in the novellas.

In fact, very little of the Secondary Phase makes it through to the novellas, and what is carried over appears in only fragmentary form and a different context. It is possible, indeed, that the direction of travel was from the other side; that material already being prepared for the novellas was hastily used in the second set of radio broadcasts, for which Adams certainly had little time and perhaps little interest. Many listeners will agree that they are, by comparison with their predecessors, perfunctory, mechanical, and lacking in the density of engagement that is so obvious in the first six broadcasts. The second series forms an inconsequential appendix, poorly integrated with the preceding material, and a narrative dead-end.

Adams’ attention was now entirely on the novellas, which were an attempt to rework the first broadcast series, extending it, giving it an overall science-fiction narrative (the sadly unconvincing search for the true source of power in the universe), and, as noted above, purging material contributed by his former collaborator. Adams did not expect these books to spawn more novellas, and he attempted to close the second, The Restaurant at the End of the Universe, and the overall story in much the same way that the original radio series had closed, with Arthur and Ford resigned to life marooned on prehistoric earth. There are of course important differences. In the novella Zaphod and Trillian are not eaten alive by a Hagunennon temporarily taking the shape of a Bugblatter Beast of Traal, that denouement being one of Lloyd’s passages. Instead, they exit the story to an indeterminate future. Marvin, the morose robot, does indeed “die”, but in the novella he is destroyed as the sole passenger of a stunt ship deliberately crashed into a star as part of a rock concert, a ship on which he has to remain to operate the teleport for Arthur, Ford, Zaphod and Trillian as they make their escape.

Nevertheless, and in spite of the differences, the two novellas are in most respects a reworking of the first broadcast series, and they form a complete and satisfactory movement, though to my mind markedly inferior to the original six radio broadcasts.

The third novella, Life, the Universe and Everything, was not published until 1982, and is a wholly new departure with another and most likely a commercial rather than an aesthetic or narratological aetiology. The fourth and fifth books, So Long and thanks for All the Fish (1984), and Mostly Harmless (1992) are similarly desperate in character, and indeed the last of these contains coded indications that Adams was heartily sick of these further accretions, since they distracted him from other work, and that he wished to make sure that there would be no more of them. A careful reader will detect the continuing reverberations of their tortured and unhappy composition, as well as other dissatisfactions, and they are consequently depressing to read. It is hard to recommended them.

Adams failed to make a success of this shift to extended prose fiction because his grasp of narrative design was not strong, and on only one occasion, the initial radio series, produced an elegant and wholly satisfying pattern of any dimension. The first series worked because it was an anti-type of narrative science fiction, requiring only token obedience to the genre’s requirements, much as Sterne’s Tristram Shandy is an anti-type of the eighteenth century novel, resembling it in superficial points but being in substance a work of a radically different kind. To this we can add the observation that the first incarnation of Hitchhikers' was in dialogue for radio broadcast, a medium which Adams understood through long acquaintance and whose limited structural extent he could control in every detail, including the sound tracks played throughout and sometimes at nearly homeopathic but nevertheless highly effective levels of audibility. Most importantly of all, the brilliant innovation of the voice of the electronic book itself provided a device by which the drama could be suspended for the introduction of digressions that have only subtle and sometimes circuitous relevance, a further point of resemblance with Sterne.

Had that initial series not been a big hit, Adams would almost certainly have continued to work on radio and television scripts in the same manner and probably with real distinction. The episodes of Dr Who with which his name is associated, “City of Death” and “Pirate Planet”, are outstanding examples of the kind, arguably superior in practically every way to their neighbours both near and far.

However, the unexpectedly overwhelming success of the six episodes broadcast in 1978 opened up the potential for larger financial rewards and so drew Adams towards narrative fiction proper. But novels required, firstly, plot articulation skills that he did not possess, and, secondly, the development of extended passages of explanatory and descriptive prose at which he was no more than competent. This was not home ground. Furthermore, though there was much dialogue, this had to be silently absorbed by the reader, rather than delivered by carefully briefed actors and spoken against a rigorously defined aural context as it had been in the radio broadcasts. By comparison with radio drama, the printed page implies a loss of control over fine details which even the most gifted author will struggle to supply by other means.

Douglas Adams was an exceptional aphorist and micro-dramatist who was drawn against the grain of his talent into writing mass-market fiction. The transition was painfully awkward, and, crucially, the miraculously effective device of the speaking Guide does not work well in print, where it loses differentiation, and becomes just one more, insufficiently distinguished, block of speech text. There was nothing that Adams could do about this, and he was not sufficiently comfortable with prose fiction to find compensating mechanisms. Indeed, I suspect that Adams’ well-known enthusiasm for Wodehouse is an indication that he was aware of his own limitations and was much in awe of a humourist who could simultaneously achieve plausible character and comic narrative extended over many hundreds of pages, something that he evidently found all but impossible.

Adams’ verbal ingenuity and striking capacity for the invention of memorable static tableau, the ironically profound planetary nano-histories for example, were never better realised than in the first six radio broadcasts. These gifts remained and are evident in everything he wrote subsequently, but the novella was a much less adequate sabot for their payloads, which belonged in and were partly inspired by the short and formulaic comedy sketch. That they had once found an almost perfect setting in just under three hours of heavily engineered radio must be regarded as a cultural singularity. No one, not even Adams himself, has been able to repeat the formula with satisfactory let alone comparable results.