The authors report a small computer system that performs analysis of eye movements. The system might be useful in helping with a few ophthalmic procedures. It is difficult to see how the system could be of general use in human-machine studies. It is too narrow in scope and too slow in processing time. There is some evidence that the authors can detect foveation. They say that other systems cannot detect rotational movement and that they plan to. They do not claim that they can yet detect rotational movement. It is unclear whether this system is any better than other systems for detecting eye movements. The reported system is slow; they hope to speed it up. This has a familiar ring of hope. We need evidence of operating results. This seems to be a preliminary report. It would have been better if the authors had waited until they had completed further studies and refined the system before reporting. It appears that the paper is a translation from Japanese into English, with several awkward phrases. It is suggested that any interested readers contact the authors for more complete information.