A rabbit was immunized with rabbit immunoglobulins of a different allotype. The anti‐allotypic antibodies produced by this rabbit were used to immunize a second rabbit which produced anti‐idiotypic antibodies. To explain the occurrence, among these anti‐idiotypic antibodies, of “internal images” of the original immunizing allotype, a restricted and a more general hypothesis are developed. The first assumes that B‐cells can be triggered when idiotopes on their receptor molecules are recognized by the paratopes of the immunizing antibody. The second denies the existence of a specially constructed combining site on the variable domain of an antibody molecule.

Although quite a lot is known about the genetic structure of the polytypic species Mus musculus at the periphery of its range, the centre of origin and dispersion of the species remains unknown. To investigate the amount of genetic subdivision that occurs in the central parts of its range, we analysed the genetic variation in four new samples of mice coming from Iran, Pakistan, northern and southern India using 3.5 autosomal protein loci and restriction fragment length polymorphisms of three genes of the Vg gene complex of the immune system. The variation was then compared with that found in the subspecies occupying the peripheral regions of the species range. The two samples from the northern part of the Indian subcontinent were shown to be more heterozygous than the samples from any of the other regions. They also contain the majority of the alleles that exist in the differentiated subspecies at the periphery of the species range. A neighbour-joining analysis on Nei's genetic distances and a factorial analysis of correspondences on the allelic composition of each sample both place the Pakistani and Indian populations in a phylogenetically and genetically central position compared to the peripheral subspecies. These results suggest that the populations in this geographically central area have retained most of the ancestral polymorphisms, which in turn indicates that the Indian subcontinent is probably the cradle of the species. The nature of the genetic relationships between the various populations throughout the species range and the possibility that they form an incipient ring species are also discussed. Our results are in agreement with the classical model of geographic differentiation where genetic divergence in allopatry is considered to be the prime cause of subspecies formation that may eventually lead to partial reproductive isolation on secondary contact.

Gene conversion by the corresponding gamma 2b gene has been proposed to explain the multiple differences between the nucleic acid sequences of BALB/c (Igh‐1a) and C57BL/6 (Igh‐1b) gamma 2a immunoglobulin allelic genes. However, genetic analysis indicates that duplicated forms of gamma 2a genes are not only present in Eastern Asia, but also in European wild mouse populations which suggests a widespread phenomenon. In order to verify whether the gamma 2a‐related isotypic genes, namely gamma 2c and gamma 2a, could correspond to those present as alleles in domestic mice (Igh‐1b and Igh‐1a), a genomic library from Mus m.musculus strain (MAI) was constructed. Extensive mapping of the recombinant phages and Southern blot analysis with several restriction enzymes gave the complete organization of these loci: gamma 2b (18 kb) gamma 2c (17 kb) gamma 2a (14 kb) epsilon. The homology in flanking, coding and intervening region sequences indicates that MAI gamma 2c and gamma 2a related genes correspond to C57BL/6 and BALB/c Igh‐1 alleles respectively. Also, Southern blot analysis using several probes derived from exonic and intronic regions between gamma 2b and gamma 2a genes shows a 2.0‐ to 3.0‐kb difference in the distance between gamma 2b and gamma 2a genes of BALB/c strain as compared to C57BL/6. Taken together, these results indicate that BALB/c and C57BL/6 gamma 2a genes could originate from different isotypes.

Based on the analysis of V alpha gene segment deletions in a panel of T lymphomas, we have constructed a map of the mouse T cell receptor alpha/delta region and assigned the relative position of 72 distinct V gene segments. Three major observations have emerged from such studies. First, members of a given V alpha subfamily are not organized in discrete units along the chromosome but largely interspersed with members of other V alpha subfamilies. Second, analysis of the deletion map suggests the existence of repetitive patterns (V alpha clusters) in the chromosomal distribution of the V alpha gene segments. Third, the present‐day organization of the V alpha/delta region may be readily explained by a series of sequential duplications involving three ancestral V alpha clusters. Direct evidence for the existence of these unique structural features has been gained by cloning approximately 370 kb of DNA and positioning 26 distinct V alpha gene segments belonging to six different subfamilies. Finally, the relationships existing between the V alpha/delta gene segment organization and usage are discussed in terms of position‐dependent models.