memberships at reasonable subscriptarwi prices, in analogy to the way institutional income from the primary jour mais allows All and its societies to provide members with primary services at low subscription prices.

Other abstracting and indexing ser vices, such as Chemical Abstracts Ser. vice, Engineering Index Bulletin Sigmaletique and Referatuni Zhurnal, should recognize that we encourage them to use the abstracts from AIP and society copyrighted journals, so long as they do not produce Englishlanguage, secondary services in the scence of physics that detract from Dervices AIP could produce for its own Dociety memberships with its own copy. Fighted material. Therefore, we plan to institute procedures for licensing the use of our copyrighted material by other services, and we expect that in most cases these licenses will be readily granted.

The services offered by the National Landing Library, Boston Spa, UK, provide me with my third example of new copying techniques that affect our operations here at AIP. This library has developed an overnight mail service through which copies of articles from any journal can be supplied to customers in the UK at low cost. Such an operation is the forerunner of future Similar services in every major country of the world. The major English-lan genge abstracting and indexing service in the science of biology is eagerly awaiting the arrival of such services in

the US, and in that same field an investigation is in progress "to discover whether there is not a large number of journals for which one copy could adequately serve US, British and Canadi. an users."

The position AIP and its societies takes on developments such as these, intended to provide better access services to the journals, is, of course, fa vorable. Indeed, we are eager to see such services growing, and plan to sup port them with the products and ser. vices that are their raw materials. However, just as with the Soviet photocopies and the use of copyrighted ab. stracts mentioned earlier, we should be recompensed for subscriptions lost be cause of these services if we are to maintain financial viability.

The financial situation

How much money is involved in lost subscriptions from, say, just the three examples cited above?

From the information supplied by the USSR on their photocopied journal sales, we know we have lost $300 000 each year from that category alone. Add another estimated $100 000 for losses due to their translation journals and book collections made up of AIP. published articles, and we find a total loss to AIP from the Soviet operations of more than $400 000 per year. For the loss of income to AIP and societies resulting from the lack of a licensing agreement with IEE for Physics Abstracts we can look at the conclusions

of IEF's negotiating team as they were stated during the summer of 1973. That team agreed with the concept of AIP receiving $190 000 per year for the use of the computer tape, and we can therefore assume this to be a minimum estimate of the annual loss in AIP income from this source.

My third example above, cover-tocover reproduction of articles from AIP and society journals, gives rise to a loss of income that is much harder to figure than it was for the first two examples. The loss of subscriptions that the AIP and its member societies have suffered over the last five years has been substantial-see figure 1. We have lost about 20% of the total number of subscriptions we had in 1966. Domestic non-member and member subscription losses account for most of this decline, while foreign subscriptions and total society membership have been relatively stable. We therefore make the assumption that the subscription loss is attributable largely to wholesale copying of single articles by institutions in the US. An estimate of the dollar value of the subscriptions lost for this reason is about $400 000 per year.

The total estimated losses for these three effects is thus about $1 million per year, an estimate that is admittedly crude. If this money were available to AIP and the member societies, page charges to authors and subscription prices to readers could both be decreased, with obvious benefits for the

"advancement and diffusion of the knowledge of physics." Copyright principles

To protect the financial viability of AIP and society publishing operations, the appropriate principles involved in the copyright process need to be presented and understood in some detail. We should remember that the particulars of the copyright claimed for a given physics article will affect four different individuals or groups: the author of the article, his employer, the publisher of the journal and individual users and republishers. Let us consider each in turn.

First, the author of the article. He originates the material that is published and usually makes the decision on where it should first appear (with either active or tacit agreement of his employer). He decides whether the subject matter of the article should be patented, whether it should be supplied to a publisher for a fee, or to a publisher (such as AIP) who expects a fee in the form of page charges. Once he makes this decision, the author should comply with the conditions of the publisher-normally stated in the journal or by a separate letter.

If the author decides to publish with AIP or one of the member societies, he should be asked by the publisher to assign, in writing, full publication and republication rights to the publisher. This request should be made at the time the editor accepts the manuscript for publication.

Should the article be subsequently republished, either by itself or as part of a collection of articles, the original publisher should give the author the opportunity to register errata or corrections to the material as firet puolished. Thereafter, the publisher should serve for the articles in his trust as the signtific and financial negotiator with republishers.

The author should have the right to make nonprofit or noncommercial ure of his work, provided he affixes to each copy, in the position legally required, the copyright notice used by the AIP or Society publisher when the article was first published. To make or authorize commercial use, for profit, of his work the author must first obtain the written consent of the AIP or society.

I mentioned earlier that the copy. right protection is limited to the expression of ideas in the published work and protects against outright copying of the work but not against copying of the ideas. The author must be aware of this limitation.

The author should be given the opportunity to write his own abstract, and also to assign the appropriate clas sification and indexing terms required when the article is inserted into the

data base of an information-retrieval system. He can thus be assured of the scientific validity of both abstract and classification. This author-written ab. stract then becomes as much a part of the copyrighted material of his article as the individual paragraphs of the main text, or the figures, tables, and so on."

The second individual affected by copyrighting procedures is the author's employer. He normally helps to defray the cost of publication of his employee's manuscript. If it is to appear in an AP or member-society jour nal, by making a page charge contribution, in retum he may want to retain some rights over the published materi al. For example, the employer might give permission for first publication only, reserving to himself the nights for republication if such conditions are accepted by the publisher. If no conditions are stated when the manuscript is submitted, the publisher must assume that there are none except those dictated by custom or tradition.

When the employ: is the US Govemment we have a special case. Articles

written by US Government employees as part of their official duties are in the public domain and are not covered by copyright.

The publisher is the third individual for whom copyright interpretation is important. He may elect to publish only that material for which he has full publication rights, both for initial and republication AIP-owned journals operate under the principle that unless otherwise stated, submission o. a man. uscript is a representation that it has not been copyrighted, published, or currently submitted for publication elsewhere.

When a publisher such as the AIP or a member society copyrights en issue of one of its journals, the nights apply to the whole issue. Such a copyright gives the publisher, as against third parties, "the same rights as if he had secured a separate copyright on each individual piece. *** This statement applies equally to the copyright protec

tion of each individual abstract of a copyrighted issue of Physics Abstracts, for example, as it does to the protec tion of each individual abstract in an AIP or society copyright journal For this reason, abstract services (such as Chemical Abstracts Service) have LDsisted that their copyright notice appear on copies made, under license and for a fee, of pages and abstracts from their abstracts journals even when these abstracts are taken verbatim from copyrighted journals.

Lastly, we should consider the rights of individual users and republishers. An individual scientist has an accepted right to copy a copyrighted art cle for his own use under the traditional copyright concept of "fair use" Republishers, as in the three examples in this article, will at times use the argument that they are operating under the "fair use" concept. Clearly, when a republisher uses every page or every abstract in a systematic. production manner, whether for commercial or noncommercial purposes, he is doing something more than "fair use" and is in infringement of copyright if he does so without permission of the copyright owner.

What are the prospects?

more

Our examination of the copyright issues that face AIP and its member societics has shown how complex are the problems that arise under US copyright laws, then how much complex must be the international implications! The exampies quoted earlier in this article demonstrate in some degree how the AIP and member societies publishing program interacts with the programa of foreign publishers, libraries, and ao on, each operating under the copyright law of his own country. These other nations too are taking a hard look at copyright leg siation in the light of modern developments, with the result that we can expect a shifting pattern of interrelating national copyright laws to affect our physics journals for some time to come

One example of the kind of change we might expect ts the licensing scheme, varieties of which are being tried out in at least three countries Sweden. France and Canada. The Swedish scheme permits multicopying of works protected by Swedish copyright only on payment of a small fee. Surveys indicate that 150 million page-copies are made in Sweden each year, rough estimates for the US suggest that several billion page copies are made here per year. Even if the new scheme works in Sweden (and it is still too new for conclusions to be drawn) we cannot be sure that a similar plan would be appropriate here

While we are moratorine possible US developments in the national copyright