There are two typical formulations for the mobilities of jumping units in concentrated polymer solutions on the basis of free volume concept. One is due to Fujita and the other to Vrentas and Duda. The prevailing notion that the former is contained as a special case of the latter is not correct, because the two theories are built on significantly different assumptions. The present paper clarifies this point and gives new evidence for the Fujita-Kishimoto approximation to the composition dependence of the fractional free volume.