My friend Arthur and I began building a website at www.scienceclassics.org that would have contained scanned copies of classic scientific papers available for free. Examples of such papers include Watson and Crick's discovery of the double helix structure of DNA, Chadwick's discovery of the neutron, and Millikan's measurement of the electron's charge. We believe that many scientists would enjoy getting an historical perspective on the development of scientific knowledge, and would appreciate having many of the papers in one location. It also seems that all biologists (and probably all scientists) should have read Watson and Crick's paper, especially since it is so short and easy to read.
Unfortunately, we found that the copyright for scientific papers often extends for 95 years, meaning that we would only be able to include papers published before 1905. This is a serious problem because scientific papers before this time were often in book format and not in English. We attempted to get permission from some journals to reproduce a few of their papers but we found that many were quite strict. In addition, it looks like at least one of the best journals (Physical Review Letters) is planning on scanning all of its past publications within the next few years, meaning that our work would not be relevant for too long.
I have recently found a website with a much more ambitious plan (i.e. with much more finanial backing). The organization is called JSTOR, presumably for Journal STORage. They have scanned in the entire text of dozens of journals, although their science collection is rather spotty. To illustrate this, there is one section for botany and another section called "General Science," which includes all of physics, chemistry, and biology. Nevertheless, the collection is beautiful and it is often fun to just roam hrough their collection.
There is also a nice website maintained by MIT (classics.mit.edu) with the text of almost 500 classic historical documents (in this case "classic" means ancient). They have the text of Herodotus' The History, Thucydides' The Peloponessian War, Euclid's The Elements, and many others. It is a wonderful collection of work and I recommend checking it out if you have any interest in history.
Unfortunately, we found that the copyright for scientific papers often extends for 95 years, meaning that we would only be able to include papers published before 1905. This is a serious problem because scientific papers before this time were often in book format and not in English. We attempted to get permission from some journals to reproduce a few of their papers but we found that many were quite strict. In addition, it looks like at least one of the best journals (Physical Review Letters) is planning on scanning all of its past publications within the next few years, meaning that our work would not be relevant for too long. 