PARENTAL CAKE AMONG FRESH-WATER FISHES. 461 



fish's tactile unil fringe. The worm, once discovered, disappears in a flash down 

 tlie catfish's tin-oat by suction and is usually not seen again, although soiue- 

 limes it is brought up to be partially masticated. On the other band, the pres- 

 ence of a young trout in the tank is not. it would seem, detected in the same way. 

 or so quickly as a worm or a fragment of liver. It more often discovers itself 

 to the Muhtptcnirus, by coming in contact with its body, when it is instantly 

 liartially disabled by an electric discharge, and pounced upon, when it likewise 

 disappears in the same manner as the worm. 



Another of their chciracteristics adverted to by P'orbes is a spirit 

 of pugnacity carried to an extreme which renders their assemblage 

 in an aquarium dangerous to some of the individuals and require 

 them " to be kept isolated in separate tanks." The individuals in the 

 atpiaria of the Egyptian Zoological Gardens were comparatively 

 peaceable. According to the director (1905, p. 27), " besides the two 

 Malapteruri " kept in one tank were " several ' bolti,' Tilapia nilo- 

 ticar and " neither sj^ecies of fish appears to in any way interfere 

 VN-ith the other." According to Forbes : 



Although supposed to be immune to each other's electrical discharges, the fol- 

 lowing observation would seem to throw some doubt on whether this lie really 

 the case. On one occasion the partition between two halves of a tank, each 

 (ontaining a Malapieninis, becoming faulty, two strong and healthy fishes 

 managed to get together, and were found by night watchman on his hourly 

 rounds, fighting with each other. Adjusting the partition, he returned the com- 

 batants to their separate cells. On his return, however, an hour later he 

 found the barrier had again slipped, and both fishes were in the same com- 

 jiartment, but one was dead. On examination no external marks of violence 

 were visible, and we can only suppose that the stronger fish had killed its 

 neighbor by a powerful electrical shock. 



The raad is apt to kill oti' other fishes of corresponding size that 

 encroach on its territory, although this is not always the case. It may 

 also make use of them in a way which reminds one of the attacks of 

 jaegers on gulls and their consequences. One was confined in an 

 aquarimn in Liverpool with a Clariid catfish ; it woidd not eat worms 

 offered to it directly, but after the Clariid had taken them, it Avotild 

 give its companion a shock which had the result of causing him to 

 vomit the worms recently taken, and these the raad Avould appropriate 

 for his own use with an open countenance. 



Sounds, described as being " not unlike the hissing of a cat," were 

 heard to issue from a raad in captivity by a sister of AVilliam Sriren- 

 sen, when staying at Mansourah on the Nile. '* This appeared some- 

 what strange" to Sorensen, but after he had dissected the fish, he 

 " coidd easily account for it by the long, narroAv ducts through which 

 the air has to pass from the anterior to the posterior cliamber of the 

 air bladder" (1895). 



The manner of reproduction of the raad is still unknown. The 

 eggs are rather large, having a diameter of a couple of millimeters, 



