Report on the Hj'droids. 



355 



espèces à stolons se rencontrant généralement snr le littoral", a con- 

 dition that Hartlaub also points ont. 



P. Cerfontaine 1902 (19) remarks, that ends of branches of Pen- 

 naria Cavolinii can be transformed to stolons, and stolonized ends 

 from two branches may form anastomoses. Gast & Godlewski 1908 

 (22) have experimented with the amputation of hydranths and 

 branches of Pennaria Cavolinii; from the surface of the wound new 

 hydranths grew out or, if the surface of the wound was in touch 

 with the wall of the aquarium or any other solid object, a stolon. 

 If the outgrowing coenosarc on a regenerating branch came in 

 touch with a solid object, it fastened itself thereon and formed a 

 stolon (p. 83). 



I may also mention N. M. Stevens' experiments, 1902 (53), on 

 Antenmilaria ramosa. Miss Stevens cut the colony into pieces, placed 

 the pieces in ditferent ways in the aquarium to see how they would 

 regenerate, and found among other things: "Certain parts of the 

 stalk tend to produce roots, others stems. Basal pieces usually pro- 

 duce stems; median pieces roots; and apical pieces, cut within the 



region of growth, tend to continue the stem. The coenosarc 



of Antennularia ramosa is of such an inditferent character, that it 

 may be withdrawn from one form of growth and put forth in an- 

 other form without the production of new tissue". 



Of greater interest than these phenomena of regeneration under 

 abnormal circumstances, is the occurrence of stolonized branches in 

 nature, and examination of the conditions that lead to their formation. 

 At present I am not able to discuss this subject thoroughly; as said 

 above, I only wish to state here the actual observations I have made. 

 These embrace partly bridges and coal- 

 escenses on polysiphonic branches, partly 

 transformation of monosi phonic branches 

 to tendrils and stolons. A polysiphonic 

 branch is in reality a collection of rhizo- 

 cauli or stolons creeping on one another 

 instead of on a solid support, hi polysi- 

 phonic species where the peripheral tubes 

 as well as the central may bear hydranths, 

 these tubes are to be considered as rhizo- 

 cauli analogous to, for example, the 



creeping stem of a Calijcella; new tubes are continually developed, 

 which creep along the stem and the branches and cause the growth 

 in thickness of these, but they can also produce a binding together 

 of several branches of the colony. When two compound branches 

 cross each other, some of the new, growing tubes may leave their 

 XLV. 26 



