Both octonion_test.cpp and quaternion_test.cpp, at about line 94, contain the following: // Provide standard floating point abs() overloads for MSVC //#if BOOST_WORKAROUND(BOOST_MSVC, < 1300) || (defined(_MSC_EXTENSIONS) && BOOST_MSVC < 1310) That looks like a workaround for the fact that prior to VC++ 7.1, the Microsoft C library did not define abs(float) etc. if _MSC_EXTENSIONS was defined. The Intel compiler uses the Microsoft libraries, and defines _MSC_EXTENSIONS, but of course BOOST_MSVC is not defined. That causes the workaround code to be included, resulting in multiple definitions, if the Microsoft library 1310 (7.1) or later is being used. I believe the corrected code should read: // Provide standard floating point abs() overloads if older Microsoft // library is used with _MSC_EXTENSIONS defined. This code also works // for the Intel compiler using the Microsoft library. #if defined(_MSC_EXTENSIONS) && defined(_MSC_VER) && _MSC_VER < 1310 I've committed this change to CVS for both source files on the assumption that this suits the library maintainer. If not, or there is something wrong with the fix, please feel free to revert the changes. --Beman