Here's a nice TED talk on the astrolabe, thanks to Alun Salt for the tip-off. The speaker uses an example of an astrolabe from the Oxford Museum for the History of Science (also featured in Alun's blogpost). The Science Museum have some pretty gorgeous ones too.
I won't rehearse what an astrolabe is here, watch the video. But I can use it to say something about children's science books. The first manual for the Astrolabe was written for a kid (Geoffrey Chaucer's son Lewis, yes that Chaucer). The British Museum has an astrolabe they think matches the one the Chaucers would have used. This book is often described as first children's book. So, the first ever children's book was a science book.
This little fact-ette pleases me immensely. Obviously it relies on a rather ridiculous (not to mention anachronistic) over-simplification of our definitions of "children" "science" and "book". I don't care though. When people at children's literature studies conferences look at me with incredulity when I say I study science books (people have, quite seriously, looked down their noses and informed me "but, non-fiction isn't literature"), I love to direct them to Chaucer.
Via Peter Hunt (1994) An Introduction to Children’s Literature (Opus, Oxford: pp.189) if you want a full bibliographic reference from a professor of children's literature studies.