A STUDY OF RESTITUTION MASSES. 459 



fragments, which might be portions of larger cells. This may very 

 possibly result from pressure through such a small mesh as 180. This 

 experiment gave no further results. 



Experiment 4. — 14th December. Squeezed a colony of Antennularia 

 antennina that had lain for some weeks in a wooden tank in the Labora- 

 tory, through 50 mesh. No attempt at aggregation. Very probably 

 the material had deteriorated. 



Experiment 5. — 19th December, 1911. Colonies of Antennularia 

 antennina brought in from the Sound. Very dirty, and placed for the 

 night in Berkefeld water. On 20th noon squeezed through 50-mesh bolting 

 silk into watch-glasses. By 4.30 p.m. small masses had formed, mostly 

 vertical to the bottom of the glass, but so far the cells not very coherent. 

 By noon on the 21st numerous small masses surrounded with perisarc 

 adherent to the glass. 



Many of these masses gradually died oi!, without showing any marked 

 change in shape. The contents, however, gradually contracted away 

 from the perisarc, and finally died. Two small masses were alive on 

 20th January, 1912, 34 days from the commencement of the experiment. 

 These were fixed for sectioning. 



Experiment 8. — 28th December. Colony of Antennularia ramosa 

 squeezed through 50 mesh at 5 p.m. into outside water. By 11 a.m., 

 29th, many large masses formed, and adhering to the glass. On 30th the 

 perisarc clearly defined, and the interior plasm slightly contracted away 

 from it. These masses lived until 5th January, 1912, when they died. 



Experiment 10. — 12th January, 1912, 4 p.m. A very fine colony of 

 Antennularia ramosa squeezed through 50 mesh into watch-glasses which 

 were placed in bowls of outside water. A larger amount of material was 

 used in this and subsequent cultures. 



By 13th three large masses, between 3 and 4 mm. in length, had formed. 

 They had not adhered to the glass, and the edges were rounded and 

 turned over, somewhat resembling the helix of the human ear. The 

 surface smooth, showing that perisarc had formed, and colour the charac- 

 teristic yellowish green. 



As time went on there was considerable contraction, and the edges 

 became thicker. The cells appeared to migrate from the centre towards 

 the edges, so leaving a thin central nearly clear plate, surrounded by 

 thickened ridges. There was no sign of proliferation or budding of any 

 kind, and by the 24th all three had lost the yellow colour, and looked 

 grey and unhealthy. They were then fixed. 



