Once the procedures for an adoption had been completed, an adoptee was expected to employ the nomenclature of the adopter. An example of this has already been reviewed, where Cicero, in a letter congratulating Atticus on his testamentary adoption, heads his letter with the revised nomenclature: Q. Caecilius Q. f. Pomponianus Atticus (Att. 3.20, dated 6 October 58 bc). The lack of a tribal indication in the Ciceronian formula is noteworthy. Cornelius Nepos independently records that the adoption was testamentary (Nepos, Att. 5). It is uncertain whether Atticus actually employed this handle, since Cicero's mock formalism seems to be intended simply to congratulate Atticus on taking up his inheritance from Caecilius. Caecilius was a knight and had no cognomen (Doer [1937] 82). After this time Atticus' freedmen were Caecilii, and his daughter was known as Caecilia Attica (Salomies [1992] 8). Atticus may well have taken on the revised nomenclature. In the case of M. Gallius, who adopted Tiberius in his will, the condicio nominis ferendi proved unenforceable. Political factors were influential. In cases where an adoptio or adrogatio is known or surmised to have taken place, changes in nomenclature are often observable, but formal requirements are unknown, and the picture is far from complete. It is assumed that formal adoptions inter vivos resulted in a stronger compulsion to adjust nomenclature. Other uncertain issues include whether an adoptee on emancipation retained his adoptive nomenclature, as seems inherently probable.