of each spine Is attached a well-developed nniscle for erecting the spine, 



A detailed examination of the construction of the poison gland was madec 

 The poison glands, situated as described above, are made up of only one 

 homogeneous type of cells. Small protective cells can be recognized around 

 the circumference of the gland but there are none to be seen between the ceils 

 of the gland. The glandular cells are much larger than those around the out- 

 side of the gland, some of them having a long axis of 26 microns and a short 

 axis of 6 microns. The long axes of the cells are parallel to each other and 

 lie at right angles to the body of the spine. As Figures 1, 2, and /; show, in 

 most cases each cell (pg) extends from one side to the other of the body of 

 the gland. This extension of the cells completely across the body of the gland 

 is not seen in the poison glands of other fishes such as Trac hinus and the 

 okoze and kasago . Synanceia . Scorpaena . Pterois . Pelor . Sebastodes ^ and so 

 forth. The poison glands of these fish as described by Pawlowsky closely re- 

 semble in other points of their structure those of the aigo which I have 

 studied, but they differ in having smaller glandular cells which are all sup- 

 ported by supporting cells (Stutzzellen) . 



A consideration of the individual cells of the glands shows that their 

 nuclei (n) (Figures 2, 3, and l) are very small in proportion to the bodies of 

 the cells. As is usual with the nuclei of glandular cells, their p>rotoplasm 

 is coarser in texture than that of other types of tissue and therefore their 

 color absorptive power Is weak, staining only slightly with hematoxylin^ 



The protoplasm is abundant and presents a densely granular appearance. 

 It Is extraordinarily eo sin- positive and stains a bright red with this dye. 

 It should be noted here that there are occasionally present within the proto* 

 plasm large round globules of nonstructural ch«iract3ro Those are also eosin- 

 positive and appear to be a colloid which is quite viscous. Even after paraf^ 

 fin embedding, staining, and washing with water they maintain their outline 

 clearly, I interpret these as drops of poison. An examination of the cells 

 of the gland also reveals the presence of a comparatively large number of 

 vacuoles in the protoplasm. These vacuol«?s are rather numerous in some cases 

 and comparatively few in others. Figure 3 shows an example in which they are 

 comparatively numerous. The significance of these vacuoles is not clear, but 

 it cannot be thought that they existed as vacuoles at the time when the fish 

 was fresh. They must have been filled with some fluid which was lost d'oring 

 the microtechnique process and which is thought to have probably been i*elated 

 to the poisonous secretion. Similar vacuoles also appear abundantly around 

 the periphery of the globules. 



The body of the gland is as described above, but if we consider the 

 question of whether or not there is a secretory duct attached to the gland, 

 we must say that there is nothing which resembles such a duct. The body of 

 the gland is made up of completely homogeneous cells, and even if we assume 

 that the secretion which fills the space between the longitudinal grooves of 

 the spine and the connective tissue issues Into the inner part of the spine, 

 there is no sort of a structure provided to conduct it to the outer part of 

 the tip of the spine. Furthermore, since there is no muscle tissue around the 

 spine there is no mechanism for compressing the body of the gland internally. 

 With the body of the gland limited to the location described above, what one 

 would like to know is how the aieo makes any use of the poison which it stores 

 up. 



An attempt was made to examine a living aieo . YRien held for Inspection 

 the fish appeared to be frightened and spread ell of its fins so that the 



219 



