464 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxix. 



halfway down the poucli; and at 11.06 o'clock. These observation.s 

 were made at night, between 0.45 and 11.30 o'clock, in the brightly 

 lighted laboratory. It is very probable, however, that the transfer 

 may take place at any and all hours of the night. It is to be noted in 

 passing that the tishes seemed entirely unaffected by the lights. No 

 attempt to handle them was made. (See Lafont.) 



It does not seem likel}^ that all the eggs are transferred at once — 

 first, l)ecause of the curious means used to move them backward in 

 the poucli; in the second place, because males are frequently found 

 with the pouch only half filled; thirdly, because males with eggs of 

 two and three stages and layings are not infrequent. When the above 

 processes have been repeated several times, the animals are seemingly 

 exhausted and remain quiet for at least two and one-half hours (the 

 extent of my observations). On this same night a third small male 

 in an aciuarium with three females '•'courted" two of them alternately, 

 but no transfer was made, though they had protruding oviducts. For 

 coition to take place, it seems necessary that the tishes should be 

 nearh^ equal in size. A ripe female paired with a male three-fifths 

 her size dropped her eggs into the water. 



This curious love play above described is not without parallel in 

 other lower vertebrates. Jordan (1891) records f oi' Diemyctyhi.^ a very 

 interesting series of observations of a courtship, lasting several hours, 

 in which caressings play an important part. Dean (18',»5), in his 

 account of the spawning of Lepidosteus, describes how the males wntli 

 wide-spread tins swim around the females and caress them with their 

 snouts. Nor is such a courtship unknown among the invertebrates. 

 Racovitza (1894) has described how the male of Octopus vulgaris 

 strokes and caresses the female. All these contacts seem to be 

 intended to excite the animals preparatory' to the sexual act. 



The arrangement of eggs in the pouch depends wholly on the size 

 of the latter. There are alwaj^s two sets of eggs, one on each side. 

 Each set may consist of one, of two, or of three rows of eggs, and 

 these may be one or two eggs deep. As noted, there ma}" be one, 

 two, or even three deposits of eggs in one pouch. In what order 

 these young would emerge from the pouch I can not sa}". Ordinarily 

 the seam breaks at points all along its length to set free the j^oung. 



The age at time of hatching can be given as ten days (with a variation 

 of eight hours) from one lot only. These young lived four daA's, feed- 

 ing on copepods with the same bird-like motion of the head and the 

 same smacking mouth motion found in the parents. In another case, 

 when the father died four days after the transfer, the little tishes were 

 with free tails. 



The eggs within twenty-four hours after deposition may easily be 

 extracted from the })ouch, coming out in masses, without injuring the 

 father. In two cases, males relieved of eggs received a fresh lot during 



