28 NATURAL HISTORY OF 



resemble in structure the siraple hydra. If we imagine 

 the young hydrte produced by budding from the body 

 of a hydra to continue attached to the parent instead 

 of dropping off, and the process to continue indefinitely, 

 we obtain a good idea of the formation of a hydroid 

 colony. It will be understood that in such a case the 

 food taken in by any one hydra would serve for the 

 general nourishment, and that really the colony would 

 be one animal with many mouths, its different parts 

 being connected by extensions of the body wall. This 

 structure will be more readily perceived by reference 

 to Plate I. fig. 2. Most of the hydroida secrete 

 chitinous coverings called polyj^aries or loerisarcs, which 

 protect the soft connecting tissue, and form receptacles 

 in which the separate polypites are lodged. The 

 connecting tissue is called the coenosarc. It is hollow, 

 and leads into the digestive or body cavity of each 

 polypite. This coenosarc contains the two layers of 

 cell structure already described as forming the body 

 wall of the hydra, and is, in fact, an extension of such 

 walls between the several polypites. The coenosarc is 

 attached to the polypary by processes which are in 

 nearly constant movement. This movement is slow 

 and amoeboid, new processes of attachment being 

 developed, and fine films of the ectoderm gradually 

 growing larger and thicker, thus showing a close 

 analogy with the pseudopodial processes, by means of 

 which some rhizopods are attached to their shell.* 



In Plate I. fig. 2, the separate polypites with their 

 respective ectoderm (ec) and endoderm (en) are seen to 

 be lodged in openings in the polypary (d), which en- 

 closes the coenosarc (c) . The hollow (a) of the coenosarc 

 * Dr. A. Weismann, Zool. Anzeig. iv. (1880), 61—64. 



