Sidney Nolan's painting "Streamers"- one of his Kangaroo series

PUTTING THE WRONG ending on Kangaroo was a major blunder, particularly for the CUP edition.

The entire CUP enterprise had been predicated on the unsatisfactory nature of the text of Kangaroo, and in particular its variant endings.

As Dr Roberts, the co-general-editor of the initial editions, had pointed out in the early 1970s, when calling for a new definitive edition of all Lawrence works:

Kangaroo and Women in Love are textually complicated works, and the texts differ in the various editions. Kangaroo, I think, is more complicated than Women in Love....I don't think there is now a text of Kangaroo in print anywhere with the text he wanted.

After the CUP's Michael Black invited me in 1975 to put in a proposal to edit Kangaroo for the CUP edition, I looked into the matter of the variant texts of the novel in some detail, having been sent copies of the holograph manuscript and the three "Berg" typescripts.

So, two decades later, when in 1994 I opened my copy of Dr Steele's CUP edition of Kangaroo, I was not only interested to see what he had to say about the provenance of the plot, but also to see what choices he had made in deciding what the definitive text should be.

In particular, I wanted to see how he had handled the complex matter of the different endings.

When I was still working on my own CUP proposal, Dr Roberts had sent me an analysis of the two variant texts made in 1965 by Dr FP Jarvis, an American bibliographer. Jarvis had pointed out that there were more than 120 substantive differences between the two published texts: the British or Secker text and the American or Seltzer text.

What most concerned Dr Roberts was that - as Dr Jarvis had pointed out - the Seltzer version ended 375 words short of the longer Secker text, which, after Heinemann took over the UK edition, had become the main circulating text of the novel, not only in the British publishing area, but in America as well.

(The Seltzer edition, as well as ending 375 words short of the UK Secker edition - with the words "broken attachments, broken" - also included a non-authorial final full-stop. This non-authorial full-stop was to play a significant role in what subsequently ensued.)

Clearly, having two texts with different endings was a substantial anomaly, especially as the Seltzer or US text, unlike the Secker/Heinemann/Penguin and later the U.S. Viking editions, also had Lawrence's final proof corrections.