The status of Gilieae as a tribal name has been called into question recently. In this paper I will summarize the alternative viewpoints, state my conclusions, and present the considerations that support these conclusions. First the historical background.

In the influential treatments of the Polemoniaceae of Peter (1897) and Brand (1907), the temperate genera were grouped in a single tribe, Polemonieae. In my 1959 treatment (Grant, 1959) I subdivided the temperate genera into two tribes, retaining the name Polemonieae for one and taking up the name Gilieae of Reichenbach (1837) for the other. No other published name was available for a segregate tribe including Gilia and its relatives. There was no apparent nomenclatural problem with it at the time and I adopted it without giving further discussion.

The justification for my use in 1959 of the Reichenbach name is summarized in Grant (1998) and is further discussed here. Within the family Convolvulaceae, Reichenbach recognized three primary subdivisions: Polemoniariae, Hydroleeae, and Convolvuleae. Polemoniariae was subdivided into Phloginae, Gilieae, and Polemonieae. Convolvulaceae is specifically assigned familial status, but Reichenbach did not assign definite rank to any of the groups below that level; he called them all “Grup- pen.” Usage with regard to infrafamilial categories and their rank was not standardized in the early and mid nineteenth century. Standardization of such matters was commenced in 1867 (Paris Congress, 1867). However, the format of Reichenbach's treatment clearly implies that he intended Gilieae to be an infrafamilial group at a level corresponding to a tribe in later and modern usage. The first subordinate level of subdivision that he recognized within the family (Polemoniariae, etc.) occupies the position of a subfamily in later and modern works. Subordinate to it and in the right position for a tribe is his Gilieae. Gilieae is also in the right word form for a tribe and is given a brief Latin diagnosis. It meets the criteria for a validly published tribe except for the lack of an explicit designation of tribal rank and this is strongly implicit.

Although the rules of nomenclature are retroactive, in practice we normally treat early authors somewhat differently from modern authors. The latter are expected to dot all the i's and cross all the t's. In dealing with earlier authors we try to infer intent from context, and make allowances for minor omissions-otherwise we would lose many old names and many links to pioneering botanists of the past. It was in this spirit that I accepted Reichenbach's Gilieae as a tribal name validly published in his 1837 work.

However, James Reveal (pers. comm., 1998) and Porter and Johnson (2000) consider that the name “Tribe Gilieae Reichenbach” is not validly published because Reichenbach did not explicitly designate rank. Porter and Johnson go on to formally de scribe it themselves as a new tribe in order to validate the name.

Recent editions of the International Code of Botanical Nomenclature (Greuter et aL, 1994, 2000) as well as earlier ones deal with this problem. In the latest (2000) edition, Article 35.2 clearly states that the use of name terminations for inferring rank of suprageneric names of unassigned rank published before 1908 is not valid. The spirit expressed in other related articles suggests that the seemingly clear intent of Reichen-bach may not be sufficient to establish formally the rank of Reichenbach's infrafamilial taxa, and that Reveal's point of view on this, if not his conclusion, is correct.

I have inadvertently contributed to the problems surrounding this name by citing the name as “Gruppe Gilieae Reichenb.” (Grant, 1959, 1998). However, Reichenbach did not use the phrase “Gruppe Gilieae/⁵ He used “Gruppen” in a generic sense for the various infrafamilial categories in at least two ranks.

Nevertheless, Reichenbach's name Gilieae is definitely a validly published although rankless name (Article 35.3 of the 2000 Code) and thus can serve as the bas- ionym for a name at the tribal level. Careful reading of the Code reveals no reason why my treatment of Gilieae as a tribe in my 1959 work cannot be treated as the valid transfer of Reichenbach's name to the tribal level, since the basionym and its place of publication are cited and the intent to treat the name at the tribal level is explicitly stated. Indeed, Example 4 under Article 35.3 of the 2000 Code is a quite similar situation, differing only in the infrageneric rather than suprageneric position of the name involved. Thus, Reichenbach's tribe Gilieae is validly published, dating from 1959. This cannot be treated as a “Rchb. in Grant” situation, as I did in my 1998 paper, but rather is correctly treated as follows: