Save Holder (1904), previously referred to, the only other investi gator, in all the literature of oral gestation known to me, who has found an incubating male to feed, is von Ihering. He (1888) speaks of catching females of Arius commersonii on the hook, but says that males whose mouths are filled with eggs do not take the bait except on rare occasions. Excepting these authors, all writers agree that fishes practicing buccal incubation fast during this period. Such is my conclusion arrived at after 6 years' study and research on the gaff-topsail. In this fish the incubatory period has not been defi nitely worked out since it has been found impossible to carry early eggs through hatching to the stage of independent life. However, so far as it has been made out by piecing together my observations, incubation continues about 70 days.
Many authors have noted that the incubating fish at or toward the close of gestation are thin, emaciated, worn out with their labors. Unfortunately I have taken but few gaff-topsails in late incubation, and none towards its close, and have noted no striking emaciation in the fishes captured. A few, however, have been rather thin and in poor condition, and moreover it has always been noticed that incu bating males taken in June and July do not seem to have much stamina. After being relieved of their eggs they die very quickly, whereas females and non-breeding males are very active and robust, and as a fisherman once remarked "require a good deal of killing" with a fish club before they become quiet. From this it would seem that the duty of incu bation bears rather heavily on the male.
The foregoing observations and conclusions have been based on normal fish in normal condition, but before leaving this subject it is necessary to relate and account for certain abnormal cases coming under my observation. These are not only interesting per se, but offer a probable explanation for an error of nearly 100 years' standing.
On July 25, 1907, I found in the stomach of an ovigerous male a perfect egg and a fragment of another. Neither showed the faintest trace of the action of the gastric juice. On July 15, 1908, a male was taken and, after removal of the eggs from the mouth, autopsy showed the presence in the stomach of a perfect egg plus several fragments, none of which showed any trace of digestion. In 1909, dissection revealed in the stomach of a male fish a whole perfect egg having an embryo in the same stage of development as those taken from the mouth. This egg was so sound and normal that I expected to see the little fish wriggle on its yolk. Other than this the stomach con tained nothing but a whitish ropy mucus. The small intestine was very much reduced in size, inconspicuous in appearance, and was empty but for some yellowish-green watery material. The belly was lanky, but there was no other evidence of wasting or emaciation. In 1910 another male was taken having 2 crushed eggs in his mouth and the fragments of 3 or 4 more in his stomach. Here again there was not the slightest evidence of digestion.
To the writer three possible explanations of the above phenomena offer themselves. (1) In the excitement engendered in the endeavor to escape as the bunt of the seine comes in filled with floundering fish, the male gaff-topsail may give off part of his eggs and swallow others. (2) When the fisherman grasps the fish by the mouth to prevent the loss of eggs, the fish may swallow some in its writhings. (3) Most of the fish just described had been struck on the head with a fish club to quiet them. If there were any eggs between the pharyngeals at this
time they would be crushed; further, the blow by reflex action might cause the sphincter muscles of the oesophagus to relax, whereupon both eggs and fragments might be swallowed.
In none of these cases is there any evidence of the eggs having been swallowed as food. Herein my observations and conclusions agree with the only others on record in this matter. Wyman (1859) writes: "Besides some nearly mature fcetuses [ of jarra-bakka] contained in the mouth, two or three were squeezed apparently from the stomach, but not bearing any marks of violence or of the action of the gastric fluid. It is prob able that these found their way into that cavity after death, in consequence of the relaxation of the sphincter which separates the cavities of the mouth and stomach." There are now to be described, and if possible explained, three extraordinary cases which do not come within the above category.
On May 29, 1911, a large fish with a great belly was captured. Touch showed the presence of eggs in this, but they seemed too far forward to be in the ovary. Autopsy made the matter clear. The fish was a male whose stomach contained a great mass of eggs stuck together like a bunch of grapes. These eggs, measuring 18 mm. in diameter, were perfectly fresh and with them was a large number of small eggs and empty follicles, such as are extruded when a female is spawned by pressure applied to the abdomen. Evidently they had been taken up by the male immediately after extrusion by the female and had been swallowed presumably as food. The eggs at this stage are apt to be adhesive and to clump in masses. This was found to be true in the case of the great female spawned artificially, as has been noted elsewhere.
The other two cases may be considered together. A giant fish with an enormous belly was taken on May 25, 1910. Eggs could be plainly felt, but could not be spawned. Dissection revealed that the fish was a female and from her stomach were taken 23 whole and 6 half eggs together with small eggs and partly digested fragments equal in bulk to all the others. On the following day another large fish was taken which also could not be spawned. On performing an autopsy her stomach was also found crowded with 23 full-grown and a large number of small eggs, none of which, however, showed traces of digestion. In both cases it seems probable that the eggs had been swallowed immediately after extrusion and as food; hence we may conclude that the females, unlike the brooding males, are canni balistic.
These fortunate catches seem to the writer to offer the explanation of the allegation by Hillhouse (1825) and others, especially Bleeker (1858), that certain catfishes are viviparous, the evidence being that they have seen eggs and fcetuses cut out of the bellies of catfishes of various species and genera. These are probably cases in which the eggs and foetuses had been swallowed by the fish as food.
Possibly the most unique find of contents of stomach was made June 16, 1908. A male taken on this day was found by touch to have some hard object in the stomach. Autopsy revealed a big wooden splinter or fragment some 5.25 inches long. The stomach was empty except for this and a considerable amount of mucus. No suggestion can be offered concerning the presence of this anoma lous body in the stomach of this fish.