is interesting that the heptamers perfectly match between Vhz and Jh2 but with Jh, and Jh3 there is a match with only 5 and 6 nucleotides, respectively. The heptamer seems to be the more important part within the recognition sequences because in some rearrangements sequences were involved which contained only the heptamer [19, 201. The only other perfect match within the A system would be the heptamers of Vhl and Jhz. We have not found this combination. It was also not observed among the 74 A chains screened in a recent publication which appeared while this manuscript was in preparation. Only two Vh2 Jhl combinations in addition to the previously published Vh2 JA3 combination were found among these lines [lo, 111. But perhaps not enough lines have been screened to find a Vh, Jhz rearrangement. Such a combination, if eventually found, would be a strong indication that the rearrangement is due to unequal sister chromatide exchange because only by that mechanism could a V gene segment be joined to J regardless of whether it is found 5' or 3' of it. However, if the DNA between V and J is simply deleted during the rearrangement, one might argue that a Vhl JA2 rearrangement would be impossible since only V segments upstream of a particular J could rearrange with it. The existence of VA2 Jh3 and Vh2 JA1 combinations would, on the latter assumption, position the Vh2 cluster upstream of the Vhl cluster. Since VA2 is found to be joined to all functional Jh, this system offers the possibility to measure the influence of the different introns and C gene segments on h L chain expression.