Is this done on purpose to sort of keep the barriers to entry to mathematics as high as possible?
Almost certainly not. What makes you think this might be the explanation? Unless you have further evidence, this doesn't seem even remotely plausible. As Pete L. Clark points out in a comment, it wouldn't be a terribly effective policy once the students found out, while there are plenty of other ways to increase standards if that's the goal. Furthermore, it's counterproductive: high standards are one thing, but there's no point to deliberately misleading people regarding what they need to know and then punishing them for not learning more than you asked them to. That's just not rational behavior, so if your department is actually doing this, then something has gone terribly wrong.
I find it difficult to imagine this is the case, though. Instead, I'd bet the department is trying to help students learn. Even graduate students in relatively strong math departments can find these classic books difficult to read, and they often benefit more from studying more accessible books. This may be poorly thought out if the comprehensive exams do not match the level of the suggested reading list, but it's probably well-intentioned. If I had to guess, I'd guess that the faculty assume the students should be able to solve the most difficult problems in these books, while many students assume they are doing fine if they can solve the average problems, which may be quite a bit easier. If this is the explanation, then it's certainly worth clarifying the standards.
Another possibility is that the reading list was set by a committee and is rarely changed, while the comprehensive exams are written each year by faculty members who are guided mainly by tradition and their own sense of what is appropriate, and who don't pay much attention to the details of the reading list. They might be surprised to hear that students actually took the reading list seriously. That would be dysfunctional, but it's a different sort of dysfunction than creating deliberate barriers to entry. A variant of this explanation is that the reading list may have been intended to indicate just the topics to be covered, with no implication that it should be a source of sample problems or indicate the level or difficulty of the exams.
It can't hurt to ask faculty members about this issue and explore ways in which it could be resolved. However, I'd strongly recommend against accusing anyone of doing it deliberately to create difficulties for students.