Consider Sir Robert Barker, who travelled to India in the late 1700s, and visited Benares (which is also called Varanasi and Kashi), where the local people showed him an astronomical observatory.
Mr Barker published his description in the Philosophical Transactions, LXVII, p 598, and they were reproduced in the Encyclopedia Britannica, 3rd edition, Volume 13, of the year 1797, which has been scanned by Google Books from the University of Michigan, which I have linked to, above; just click on the picture and it will take you to the full scan of the book.
By virtue of modern copyright law, items that have been published in 1797 are public domain. Actually anything before circa 1923 in the US is now public domain. Thus.
Google books, and public libraries. Google has the servers and the software and the money. The libraries have the books, and the catalog data.
Here is another version of the same image, which has apparently undergone a 19th century version of photoshopping. This version is from "Old and new astronomy", Richard Anthony Proctor and Arthur Cowper Ranyard, published by Longmans, Green and co., from 1892. http://books.google.com/books?id=KS8_AAAAYAAJ
I know it looks off-color. However if you simply download the book, save the image, and run it through a grayscale filter on any common paint program, then it will look as fine as the first image above it.
Now, consider Getty Images and it's image from the Hulton Archive.
I cannot, of course, paste the image here. But if you click on the link, you will notice that this is exactly the same picture as the picture above. The difference is that if you want to actually use the Getty Images version on your own web page, you have to pay Getty Images the sum of $338.00 for a 2 year agreement. They don't even offer an agreement longer than 2 years. Thus, they are claiming ownership. If you used it without purchasing it, you might find yourself the defendant in a lawsuit, brought by the very well payed lawyers that large companies like Getty tend to have.
The image even says it is from 1800; basic study of copyright law makes it clear that this date alone means there is no way to claim copyright of the image in the US.
It reminds one, perhaps, of other attempts by various free-market purists to claim ownership of what is clearly, by law and by custom, the property of the public. For example, there are biotech companies who have attempted to patent varieties of corn that were created by other ordinary people who were simply doing cross breeding. Then there are the 'patent troll' law firms. They have 'forum shopped' and found a district in Texas where they can sue, basically, anyone for any obvious idea, claim they invented it, and make millions of dollars.
In The Age of Turbulence, Alan Greenspan says over and over again that free markets depend on the 'rule of law' and the concept of 'property rights' being protected by that law. But what he doesn't discuss very much is that free-market entrepeneurship has had an uneasy history with the actual implementation of private property and the rule of law, from slavery to colonialism to eminent domain. When you take something that doesn't belong to you, whether it is a person, land, or in modern times, and idea, and claim it as your own, then that is stealing. It doesn't matter if you call yourself a 'developer' and you are stealing from 'the public', you are still stealing. In other words, Getty Images is stealing from us, the Public.
So, to be safe, you can always check back in Google Books to see if there is an old public domain version of an image you find in Getty, or anywhere else. Then just use the image from Google Books. But then what about poor Google? They put a lot of money into the project. But they also put a notice on all their public domain books, asking politely that people not use them for 'commercial purposes', while half-handedly acknowledging the legal truth that Google doesn't own any of those ancient books it has scanned.
On the other hand, remember where Google found all those books. They didn't buy them. They got them from libraries; libraries that your tax money payed for. And your parents tax money and your grandparents tax money. So while Google has performed a great public service, that libraries should have been doing a long time ago, it has also benefitted from the services of the public, namely those libraries.
It is a hard call. One wonders how Barnes and Noble got Google to allow all of those old books to appear on the Kindle. Did they cut a deal? Who knows. If you wanted to use one or two of those images yourself, in a movie or something, would you email google first? Even though you don't have to? Just to be polite? What if they said no? It is a conundrum.
The ironic thing here is though, that unless I had seen that Getty Images result about the 'benares observatory' as a search result, I never would have known what keywords to look for in google books to find an old picture! Which makes me a bit of a hypocrite, I suppose.
Copyright Term and Public Domain in the United States, Cornell Law School, retrieved 2011
Google Books, Library Partners page
Getty Images apparently thinks it also owns KGB documents, like this 1959 scan of defector Peter S Deryabin's identity card. Just because they lost the cold war and dissolved doesn't mean that Time Life Inc suddenly owns the Soviet Union's intellectual property. By modern law (you know, Rule of Law, as in Alan Greenspan's favorite thing to talk about when he is talking about the problems of the Soviet Union), the Russian state owns that stuff, but they have chosen to Public Domain all of their official documents. Here is the ACTUAL LAW. LAW 64629 at consultant.ru. If you are too lazy to learn Russian (like me) you can read a summary here at wikimedia commons.
Getty Images, Life Magazine, Herbert Orth, and Time & Life Magazine do not own the KGB identity card of Peter Deriabin, or anyone elses KGB identity card, just because they took a photograph of it 50 years ago and scanned it onto a website. Claiming that they do is called "Stealing".