the aanocoete squims upward in its burrov; until the oral hood- 

 ±G at or near the surface of the botton (Ficure 5U). Here it 

 may lie for lone periods of time, the branchial area expanding 

 aiid contracting as vrater is punped in and out for respiratoiy 

 and feeding purposes. Water is brought in through the oral .hood . 

 and expelled through the gill openings. Pumping action into the 

 oral hood is easily discernible by follorri.ng bits of detritus 

 suspended near the bottom as they are drarm into the hood, Ivicro- 

 scopic organisms are dra'vvn into the hood on the imter currents. 

 At least some of these organisms are separated out from the 

 detritus by the sieve apparatus (anterior to the pharynrc) and 

 passed to the intestine for digestion (Gage, 1923). Periodically, 

 the detritus accumulated on the sieve is blovm out. The larva 

 is seen to ejipand its branchial region, the gill openings close, 

 and with a rapid, convulsive movement of that region and the head, 

 a cloud of small particles is ejected from the hood. T^^ical 

 pumping is resumed at once. Occasionally an indi-iridual mil pro- 

 trude its entire branchial region from the burroiT during these 

 activities J larvae of northern brook lampreys also do this (Leach, 

 I9I1.O). At irregular intervals, the ammocoetes retreat to the 

 depths of their burrows for varying periods. 



Similar observations of larval habits in aquaria vrere made 

 by Gaga (1923). He used aTmocoetes of the sea lanprey from Cayuga 

 Lalce, New York. Our observations agree very closely. 



AquariuiPx observations mal-ce it easily understandable i^hy lam- 

 prey larvae are seldom observed in their "beds" in a stream. The 

 vibrations set up by footsteps across the floor of a wooden 

 building caused all aquarium-held specimens to i-etreat from the 

 surface into the depths of their b^'orrows. After several minutes, 

 if all remained q-^oiet, they returned to the surface again and re- 

 sumed feeding. Along a stream, footsteps on the banlc or in the 

 stream bed undoubtedly also cause the retreat of larvae. For this 

 reason, irdividuals of this life history stage are seldom seen, 

 even by careful observers, in their natural surroundings. 



Size of larvae and estimated length of larval life 



I'iy estimate of the length of the larval life of the sea lam- 

 prey is based upon analyses of the distribution of sizes of spe- 

 cimens taJcen in various seasons of the year. With the exception 

 of age-group which is positively identified, the length data 

 assigned to older age-groups is subjective. They are valid only 

 if I have correctly interpreted the age-groups present in the 

 several collections. 



Since the sea lamprey has a single spap/ming season each year, 

 the ages of the larvae in any single collection, made over a 

 shurt period of time, should vary by annual incraments. Presumably 

 then, if the ammocoete collection e:;diibits definite size classes. 



- 1U3 - 



