At present the most promising detector is a pixel type such as CCD vertex detectors[7] in order to measure dense positions of particles. We do not need to measure their energies since momenta of our interesting particles are analyzed by the 2 tesla solenoidal magnetic field as described in the previous section. The densest regions are and , whose density is estimated to be 120(particles) x 72(bunches) /4.1cm (0.32MHz/cm) if the detector can be read out only between RF pulses at 150Hz for the JLC-I[9]. The occupancy rate is below 1% in the CCD-detector with 25x25m pixel size. Assuming that the energy of electrons(positrons) traversing the detector is 200MeV, the total dose is estimated to be 110Krad for a year(100days)'s experiment. This estimated dose is very close to present life time of CCD-detectors. For experiments in several years, we should need more radiation-hard CCD detectors resisting up to 1Mrad. Since there are good prospects, megarad hard CCD detectors are available in the near future[7, 8]. In order to get any useful information inside a bunch-train of 201.6nsec (2.8ns x 72bunches), it is greatly required to have a fast-gating CCD. Since there is few possibilities to realized such a CCD in future, it is very difficult to extract informations for each bunch during collisions. However, there is a solution to get information inside the bunch-train when one of two LINACs can be operated with a single bunch which is called a test beam. This test beam can probe inside the bunch-train in the other LINAC by colliding with the specific bunch in it at IP. We can easily choose any bunch to be probed in the train by adjusting a timing of the test beam. For this test beam operation, any information concerning a bunch property can be obtained.
As clearly seen in Tab.1, where measurement errors are statistical ones corresponding to numbers of bunch collisions listed in the second column, beam sizes () and vertical displacements can be measured sufficiently even for a single bunch. For the measurements of transverse rotations a few bunch trains must be necessary for the statistical accuracy.