STARVING PUPS. 417 



Kitovi, 1,070. The total of live pups for Kitovi i.s 5,940. To this should be added 

 100 dead pups to find the total births (6,0-19) ot this rookery for the season of 189(>. 

 The cows counted on Kitovi in the height of the breeding season numbered 3,l.'i2. 



A cow with one hind Hipper bitteu ofl' squarely at the angle of the body was 

 seen. The sore was fresh; otherwise she was all right. 



One pup Jumped off a rock feet high, lighting ou his nose on a sharp stone. He 

 seemed surprised, but went off as if unhurt. Two pups leaped off in very high surf. 

 Failed to a[)i)ear anywhere. When last seen one had its mouth open panting. 

 I'robablj' both drowned. One very large cow was noticed. She weighed probably 

 100 pounds, and had white whiskers. 



Many of the pups were so full of milk that they could hardly waddle. They 

 often voided excrement when hurried. Some of the cows are looking fat, as if 

 well fed. 



STARVOG PUPS. 



The presence of starving pups is evident, some staggering along in the rear of 

 every pod driven off. When mixed up with the general herd of pups they are not so 

 conspicuous as when an effort is made to drive them. Then the starvelings fall 

 behind. Some of these seem less emaciated tlian the one we have at the house 

 for experiment ; but they will all die within a few days. They are undoubtedly 

 chargeable to pelagic sealing. Evidently many pups will die from this cause on 

 Kitovi within the next week. Probably those to die tirst are younger ones whose 

 mothers had been at sea some time before they were caught by the sealers. 



Even a small rookery like Kitovi seems like a great city when you try to count 

 the pups. 



PROPORTIO>" OF cows AND PUPS. 



It is certain from the count of live jmps that only about half of the females who 

 breed are on the rookeries at any one time from the beginning to the end of the 

 season. There are probably more ou the rookeries at the height of the season than at 

 any other time, but there is no time when all. or anywhere near all. the cows are 

 present at one time. Probably no cow leaves until after she is impregnated. Then 

 the older cows doubtless take to the sea and the rookery spreads: that is to say. 

 extends backward through the incoming of the fresh cows, who give birth to their 

 pups in the new harems formed around the idle bulls at the back of the former 

 rookery line. 



The virgins come in also and fall in with the idle bulls, so that very few, if any, 

 bulls iu the course of the season fail to get some cows. As a rule, each cow remains 

 in the harem where her pup is born, although the form of the harem and its discipline 

 relaxes as the sea.son advances. By the middle of August the cows move about much 

 as they please. But while a cow often follows her i)up, still oftener does she call and 

 wait for it to come to her. By the middle of August the pups know the whole 

 rookery and can tind their way anywhere. They can then be driven in pods and 

 handled just as bachelor seals are. 



The error made by all observers from the first has been that they supposed that 

 thei'e was a time of greatest density and compactness and that at this time virtuallj- 



