Clapel, I am not entirely certain that I understand you correctly, but you seem to be saying that Carpentier had two nearly indentical marks with the same initials and symbol: one on the the insculpation plates of the year V and which is described by Douet, the other on the plates beginning with the law of 19 Brumaire an VI and which is illustrated in Arminjon. This is not correct.
The marks described in Douet’s Tableau des Symboles (Paris: the author, 1806) are the same as those illustrated in Arminjon. The title page of Douet clearly states that his tableau contains marks “insculpés en l’administration du Departement de la Seine, en la Préfecture de Police et en l’adminstration des Monnaies, depuis la Loi du 19 Brumaire an VI, jusqu’au 30 Septembre 1806” and his preface mentions the mandatory lozenge shape. The numbers Douet assigns the marks are taken from the plates at Préfecture de Police, and are cross-referenced in Arminjon. (Unfortunately, there is a conflict regarding the number of Carpentier’s mark: Douet has it as “287” whereas Arminjon has “267”. This is doubtless a printing error, though which is correct cannot be determined without access to the plates themselves.)
However, as Arminjon notes, the marks on the two plates of the year V conserved at the Musée de la Monnaie were not lozenge-shaped (which form was not mandated until 17 Nivôse VI [6 Jan 1798]), but rather followed the form of ancien régime marks. These two plates, with close-ups of all the marks, the names of the silversmiths and the dates of registry are shown in Jacques Helft’s Nouveaux Poinçons (Nancy: Berger-Levrault, 1980), pp. 85-96. Carpentier does not have a mark registered on these plates, nor is the mark on this tabatiere on them. Again: these marks are not lozenge-shaped.
Finally, while new guarantee marks were created with the decree of 19 Brumaire an VI (9 Nov 1797) as stated in Arminjon, the new Paris assay office was not established until 15 Prairial an VI [3 Jun 1798] (see Bulletin des lois 2:205:1862), and the new punches not received and used there until nearly three weeks later, on 1 Messidor an VI [19 Jun 1798] (see Bulletin des lois 2:209:1891.) So it is entirely possible that the lozenge-shaped maker’s marks could have appeared alongside the earlier guarantee marks of the year V for theoretically at least six months.
I have no doubt that a careful, mark-by-mark study of Arminjon would reveal the maker here, but I confess that even the tedium of quarantine has not yet inspired me to undertake the task.
Hope this helps!