The Hydroids experimented on were Antennularia ramosa, and A. antennina.

These were cut in pieces and pressed through bolting silk, with the result that isolated cells and small cell aggregates were obtained, which soon aggregated together to form compact masses.

These restitution masses secreted a perisarc within from 12 to 18 hours.

Various changes in shape, and general retraction of the mass away from the perisarc occurred later, but even up to 60 days there was no sign of the regeneration of the hydranths.

The restitution masses consisted of ectoderm and endoderm cells, and in addition such structures as nematocysts, ova, and brokendown cells, all of which were subsequently absorbed and played no part in the future development. The ectoderm cells were relatively little damaged, and were embedded in a plasmodial mass formed by the endoderm cells.

A definite layer of ectoderm cells is formed on the surface, and thesecells secrete the perisarc.

Gradual aggregation and segregation of the endoderm cells from the plasmodial mass takes place ; and they form very definite tubules similar in structure to the coenosarcal tubules continuous with the enteric cavities of the normal hydranths. These tubules are embedded in a mass of ectoderm cells, they are convoluted and ramify in all directions. Many granules develop in the cytoplasm of these cells, and after about a month many of them have degenerated.

The ectodermal cells show no signs of degeneration, and the masses containing them have been kept alive for 60 days at the time of writing.

In none of the experiments was there any sign of the occurrence ofcell division