1. Eight species of Actinians (enumerated on p. 148) have been specially studied with regard to their reproductive processes.

2. Among these forms five methods of reproduction occur—longitudinal fission; laceration by tearing; laceration by constriction; production of young viviparously; deposition of ova.

3. Deposition of ova may co-exist with any other method, except perhaps in certain viviparous forms.

4. Apart from this, one method only prevails in any one of the selected species—

A. Longitudinal fission in Sagartia sphyrodeta and Diadumene luciœ.

B. Laceration by tearing in Sagartia elegans and Diadumene cincta.

C. Laceration by constriction in Sagartia lacerata.

D. Viviparity in Sagartia troglodytes and Cereus pedunculatus.

5. In Sagartia anguicoma neither viviparity nor asexual reproduction occurs.

6. The viviparous forms studied do not reproduce asexually; nor are those which reproduce asexually viviparous.

7. In the experiments and observations of the author and of W. E. Evans, no exceptions to the above statements occurred. The question of exceptions is discussed on page 161.

8. Four of the above species were selected for regeneration experiments. Pieces were cut off from the margin of the base in each (artificial laceration).

9. In Sagartia elegans (which lacerates naturally) it was found that nearly 100% of such pieces would regenerate easily; in S. lacerata the same was true in the best experiment.

10. In Sagartia troglodytes (which does not lacerate naturally) only two pieces out of 76 regenerated fully, and the regenerative ability of the species appeared to be weak. In S. anguicoma, however, the percentage of pieces which regenerated fully (and did so strongly) was very much higher, though not reaching that of S. elegans.