(v.) Phoronis architecta Andrews.—A species recently described by Andrews from Beaufort, N.C. Its distinctive features are: "the formation of isolated tubes covered by definite collections of sand grains; the presence of special prostomial organs, possibly of use in the formation of these tubes; the great development of the longitudinal muscles; the presence of a ciliated groove in the digestive tract; the apparent separation of the sexes."

(vi.) Phoronis psammophila Cori.—Found in Faro, near Messina. The tube is hyaline, and is covered by numerous grains of sand, some of considerable size. The length of the individuals is 25 to 50 mm. There are 60 to 90 tentacles. The colour is a fleshy red. A second species discovered by Haswell in Port Jackson had no points of importance to distinguish it from Ph. psammophila, except that no sand adheres to its tube and the number of tentacles is slightly greater.

In addition to the various species of Phoronis, several distinct forms of its larva, Actinotrocha, are known, and have been named without having been traced into their corresponding adult form.

The position of Phoronis in the animal kingdom has formed the matter of considerable divergence of opinion amongst the naturalists who have studied it. The earlier writers regarded Phoronis as allied to the Gephyrea, and it was for a long time classed with these animals, but placed in a separate sub-Order, the Gephyrea tubicola, which was opposed to the Gephyrea nuda, which comprised the true Gephyrea.

Caldwell referred Phoronis, the Brachiopoda, the Polyzoa, and the Gephyrea to the same type of body structure, and Lankester subsequently suggested the provisional name Podaxonia for this miscellaneous collection of animals. Lankester divided his phylum Podaxonia into three classes: (i.) the Sipunculoidea (Gephyrea), (ii.) the Brachiopoda, and (iii.) the Polyzoa. The last-named class he divided into three sections: (a) the Vermiformia, this includes the single genus Phoronis; (b) the Pterobranchia, including the forms Cephalodiscus and Rhabdopleura, whose affinities with Balanoglossus were subsequently demonstrated; and (c) the Eupolyzoa, including the forms treated as Polyzoa in the following pages.

Masterman's recent researches[[499]] on Phoronis seem to indicate that the Vermiformia, like the Pterobranchia, must in future be grouped with the Hemichordata. He finds three well-defined coelomic spaces corresponding with the epistome, the collar, and the trunk, and also representatives of the collar pores, and is further inclined to believe that structures representing the notochord exist in Actinotrocha.

Should Masterman's researches be confirmed, Phoronis will be removed from its present isolated and enigmatical position, and placed with Cephalodiscus and Rhabdopleura amongst the Hemichordata, which will be described in Vol. VII. of this work.

POLYZOA

BY

SIDNEY F. HARMER, M.A.