MECHANICAL PROTECTION AND POISONS 383 



cytoplasm of the cell. These grow into very large drops which seem 

 never to blend, but to work their way individually down through the 

 reticulum in the lumen. Comparatively little of this matter is secreted 

 and it is possibly only a part of the material which goes to make up the 

 poison used by this creature to kill its prey, the rest being a soluble fluid 

 that does not appear in the usual microscopic preparations. 



These cells extend into the duct of the gland as low, non-secreting, 

 epithelial elements, thus proving the origin of the poison cells. The 

 poison cells of the scorpion and allied forms appear in the tail and other 

 parts of the body, but have much the same histological and cytological 

 structure as those we have just examined. 



The vertebrate animals are not without their poison-producing mem- 

 bers, and some of these are among the most dangerous creatures known, 

 on account of their size and the amount of venom that they are able to 

 inject into a wound. These animals also have the mechanical structures 

 of offensive protection remarkably well developed. 



Among the fishes some spines, which are mesodermal bony structures, 

 are developed in connection with the fins. Most of these are not poi- 

 sonous except for the slime and dirt that are associated with them. The 

 sting of the whip ray, Dysatis, and other sting rays, makes a very ugly 

 wound. It is barbed and when broken it is renewed from a tissue nu- 

 cleus that forms new spines, slowly, all the time. 



Some fishes also have the first ray of the two pectoral and the dorsal 

 fins enlarged and barbed to use as a weapon of defense. In some cat- 

 fishes, as Schilbeodes, the pectoral spine is associated with a weak poison 

 gland (Fig. 349, A and B), which appears in two forms; an axillary 

 gland opening by a pore near the origin of the fin in most of the species, 

 and a glandular tissue placed between the skin and spine in those cat- 

 fishes which do not have barbs on the spines. 



The axillary gland is clearly an invagination of the stratified surface 

 epithelium whose cells are specialized to secrete poison. They proliferate 

 and swell up until they finally burst, and the poison is discharged from 

 the pore. As can be seen in the figure, these poison cells are modified 

 clavate cells which occur in all the skin of this and other fishes. One needs 

 but to trace the row of clavate cells (Fig. 349, A and B, c.c.) from the 

 outer epithelium around and into the poison gland to realize this fact. 



In the fin spines of such of the catfishes as have no serrations on these 

 spines, may be seen another collection of the same poison cells. Here 

 they lie between the integument and the spine. No duct is apparent, 

 and we must examine a longitudinal section of the tip of the spine before 

 the relations of the poison cells to the epidermis can be understood. 



Such a section (Fig. 349, 4) shows that the epithelium on the end of 

 the spine has been reflected as a blind invagination around the central 



