Stebbing, 1906: 273.

“Pleon well developed. Head usually produced into a deflexed rostrum. Sideplate 4 (except in Argissa) not conspicuously large, often smaller than 3rd. Eyes of various characters. Antenna 1 with accessory flagellum, peduncle longer, flagellum shorter in female than in male; antenna 2 little or not longer than antenna 1 in female, considerably longer in male. Lower lip with inner lobes. Mandible robust, but palp slight, usually with a very short 3rd joint. Maxilla 1, inner plate with several setae, outer with (so far as known) 11 spines, 2nd joint of palp long. Maxilla 2, inner plate rather the broader, fringed on inner margin. Maxillipeds normal. Gnathopods 1 and 2 feeble, slender, not very unequal, 5th joint long, 6th shorter, subchelate or simple. Peraeopods 1 and 2 slight. Peraeopod 5 usually the longest. Uropods 1 and 2, outer ramus the shorter; uropod 3, rami subequal. Telson long, cleft (except in Bruzelia).”—Stebbing, 1906.

As will be seen from the generic and specific descriptions which follow, there is considerable discrepancy between the genus Cacao and the Family Tironidae in which it has been placed. For the present, I consider it of some value to give the family description here, even though Cacao may well be transferred at some later date to a different family.

Barnard places Cacao in the Tironidae with the comment that its place here may be debated Using Stebbing's key to the families of Gammaridea (1906),

the genus keys out in the Laphystiopsidae (= Lafystiidae, cf. Barnard, 1930: 342) or possibly, depending on one's interpretation of “maxilliped normal” in one dichotomy, into the Calliopiidae where Walker (1909) provisionally placed the genus Chagosia with which this genus may be identical. (Walker gives insufficient details to clarify this point). Barnard's own key (1940), which does not include the Lafystiidae, keys out Cacao into the Ochlesidae or again, depending on interpretation, into the Calliopiidae. (In this key, Colomastigidae and Ochlesidae appear to have been accidentally transposed.) In neither key does the Tironidae appear a possible conclusion.

It would seem that in placing the genus in the Tironidae, Barnard has followed the lead given by Chevreux (1912) when he placed the closely allied Alexandrella in this family. Rather than confuse the issue by further conjectures on the family relationship at this stage, I prefer to leave Cacao, at least temporarily, in the Tironidae although it does demand a certain liberality of interpretation.