HE technique of hybridisation has been widely used in the determination of base sequence homologies. The Carnegie Institution group pioneered the application of DNA-DNA hybridisation to the comparison of sequence differences between a wide range of species (HOYER, MCCARTHY and BOLTON 1964), while differences between rat and mouse have been studied by MCLAREN and WALKER (1966). Since hybridisation between DNA: DNA and RNA: DNA are essentially similar reactions RNA-DNA hybridisation may be used to detect genetic differences, provided the RNA is sufficiently representative of the sequences we wish to compare. Recently, BRITTEN and KOHNE (1967) have shown that the DNA of higher organisms includes a substantial fraction of reiterated sequences and that the earlier DNA-DNA comparisons referred to such fractions rather than whole genomes. The interpretation of these sequences is obscure and we need further information about differences between them, especially in related species. The present paper deals with such evidence, which is based on the method of RNA-DNA hybridisation, in which the RNA is synthesised in vitro (cRNA) with the aid of DNAdependent RNA polymerase and appropriate DNA primers. Although nearest-neighbour analysis shows good agreement between cRNA and the primer DNA (WEISS and NAKAMOTO 1961), it cannot be assumed that all sequences are equally transcribed. If there are species differences in the frequency of preferred sites of transcription, this will enhance apparent genome differences, and, in some situations, this may be an advantage. The use of hybridisation between DNA and cRNA has several attractions which prompted the experiments described here. It is often difficult or excessively labourious to prepare, for DNADNA hybridisation, labelled DNA in sufficient quantity and of adequate specific activity from species which may be of particular interest for reiterated sequence comparisons. Subject to the qualifications noted above, the cRNA method is a general one and the RNA may be made with any desired specific activity. Also, convenient techniques, based on the use of membrane filters, have been developed by NYGAARD and HALL ( 1964), and by GILLESPIE and SPIEGELMAN ( 1965) for the quantitative measurement of RNA bound to DNA. Finally, the treatment of reaction mixtures with RNAase offers a fairly stringent test of complementarity. 1 This paper is dedicated to Dr. THEODOSIUS DOBZHANSKY, in recognition of his outstanding contributions to the genetics of natural populations and of evolutionary processes.