ly88.] POISON-ORGANS OF TRACHINUS, 363 



sheath, and the two points (x, x) at which the glands are connected 

 with the epidermis. It will thus be seen that the end of the spine 

 somewhat resembles in its arrangement a hypodermic syringe. 



In order to make out the more detailed structure of the various 

 parts, serial sections were taken of both opercular and dorsal organs 

 in various planes. The specimens were decalcified, stained with 

 borax-carmine, imbedded and mounted in the usual way. 



A transverse section through the line a-b in fig. 1 is represented 

 in fig. 4. The spine (op.s.) in this region broadens out on both 

 sides of the grooves, which are thus narrower here than they are 

 more posteriorly ; the dorsal and ventral glands (ffl-), enclosed in 

 their thin membranous capsules (cp.), are seen projecting into them. 

 The capsule proper is surrounded by a mass of dermal connective- 

 tissue (c.t.), except where it projects into the groove of the spine : the 

 epidermis (ep.) is seen covering the connective-tissue layer externally. 

 No special muscles are present in connexion with the glands : the 

 fibres shown in fig. 6, inserted into a slight ridge of the spine, be- 

 long to one of the opercular muscles. 



The glands, which contain no lumen, consist of a number of 

 very large rounded or irregular cells ; their contents are granular 

 and in most of thein one or more distinct nuclei can be seen, many 

 of which show indications of recent division. Around the edges of 

 the glands smaller — probably immature — cells are present here 

 and there in transverse sections through this region. The cell- 

 boundaries cannot everywhere be clearly made out, and I am inclined 

 to think that in the discharge of their secretion the cells simply burst, 

 their contents passing along the grooves amongst the other cells to the 

 exterior. 



With the exception of the smaller number of cells and the differ- 

 ent form of the spine, the structure of the posterior narrower part of 

 the organ shows no important difference from the anterior enlarged 

 part. Towards the narrow termination of the gland, however, 

 shortly in front of the apex of the spine, in which region the cells 

 are smaller, the connective-tissue lying between the gland and the 

 epidermis thins out, and the cells of the epidermis and gland become 

 continuous with one another (comp. figs. 3, a, 6, 7, ej)., c.t., x). 

 Owing to the toughness of the decalcified spine and to the looseness 

 of its connective-tissue sheath, it is exceedingly difficult to obtain 

 Satisfactory unbroken sections in this region, and I have not yet 

 Succeeded in ascertaining with certainty the manner in which the 

 secretion passes to the exterior, but am inclined to accept Byerley's 

 explanation. There can, however, be little doubt that the gland is 

 developed as an epidermic involution, the whole of which gives rise 

 to secretory cells, so that there is no marked differentiation into 

 gland and duct. 



Dorsal Orffans. — The arrangement here is precisely similar to that 

 seen in the posterior narrow part of the opercular organ. There is an 

 expanded part of the gland, which extends along both grooves of 

 each dorsal spine, showing not more than three or four cells in a 

 single transverse section through the broadest part. A connexion 



