2. Balbus, his wife, and their two sons (5, 6, 13, and 14).
3. Publius, who is himself a _pater familias_, but has no _agnati_ at all.
4. Marcus, his wife Terentia Minor, and their child Decimus (9, 10, 16). Notice that the other child, Servius (15), having been emancipated by Marcus is no longer agnate to his father, mother, or brother.
§25. Cognati, on the other hand, were what we call blood relations, no matter whether they traced their relationship through males or females, and regardless of what _potestas_ had been over them. The only barrier in the eyes of the law was loss of citizenship (§18), and even this was not always regarded. Thus, in the table last given, Gaius, Faustus, Balbus, Publius, Terentia, Terentia Minor, Titus, Tiberius, Quintus, Sextius, Servius, and Decimus are all cognates with one another. So, too, is Gaia with all her descendants mentioned. So also are Tullia, Titus, and Tiberius; Licinia, Quintus, and Sextius; Marcus, Servius, and Decimus. But husband and wife (Gaius and Gaia, Faustus and Tullia, Balbus and Licinia, Marcus and Terentia Minor) were not cognates by virtue of their marriage, though that made them agnates. In fact public opinion discountenanced the marriage of cognates within the sixth (later the fourth) degree, and persons within this degree were said to have the _ius osculi_. The degree was calculated by counting from one of the interested parties through the common ancestor to the other and may be easily understood from the table given in Smith's "Dictionary of Antiquities" under _cognati_, or the one given here (Fig. 1). Cognates did not form an organic body in the state as did the agnates (§22), but the 22d of February was set aside to commemorate the tie of blood (_cara cognatio_), and on this day presents were exchanged and family reunions probably held. It must be understood, however, that _cognatio_ gave no legal rights or claims under the Republic.
[Illustration: FIGURE 1. TABLE OF RELATIONSHIP]
§26. Adfines.--Persons connected by marriage only were called _adfines_, as a wife with her husband's cognates and he with hers. There were no formal degrees of _adfinitas_, as there were of _cognatio_. Those adfines for whom distinctive names were in common use were: _gener_, son-in-law; _nurus_, daughter-in-law; _socer_, father-in-law; _socrus_, mother-in-law; _privignus_, _privigna_, step-son, step-daughter; _ritricus_, step-father; _noverca_, step-mother. If we compare these names with the awkward compounds that do duty for them in English, we shall have additional proof of the stress laid by the Romans on family ties: two women who married brothers were called _ianitrices_, a relationship for which we do not have even a compound. The names of blood relations tell the same story: a glance at the table of cognates will show how strong the Latin is here, how weak the English. We have "uncle," "aunt," and "cousin," but between _avunculus_ and _patruus_, _matertera_ and _amita_, _patruelis_ and _consobrinus_, we can distinguish only by descriptive phrases. For _atavus_ and _tritavus_ we have merely the indefinite "forefathers." In the same way the language testifies to the headship of the father. We speak of the "mother country" and "mother tongue," but to the Roman these were _patria_ and _sermo patrius_. As the _pater_ stood to the _filius_, so stood the _patronus_ to the _cliens_, the _patricii_ to the _plebeii_, the _patres_ (=senators) to the rest of the citizens, and _Iupiter_ (Jove the Father) to the other gods of Olympus.
