ACTUAL COUNTS OF PUPS. 



109 



THE DECLINE BETWEEN 1896 AND 1897. 



As n result of the investigations of the past two seasons we are able now to sub 

 mit definite and tinal proof not only of the fact of decline, but also an approximately 

 accurate measure of its rate. 



COMPARATIVE COUNTS, 1896-97. 



During the season of 1896 a very accurate estimate of the total number of harems 

 on the two islands was made. On certain rookeries and parts of rookeries careful 

 counts of the individual cows present were made at the height of the season and on 

 approximately the same dates each year. Afterwards a thorough enumeration of the 

 live and dead pups on the same breeding grounds was made. We have had occasion 

 to criticise and revise our detailed census of 1896, but this revision does not affect the 

 actual counts for that year, which we have no occasion to alter. 



These comparative counts for the two seasons are as follows: * 



Actual co nuts, 1896-97. 



a Not counted. 



b Count of 1896 rejected as obviously incorrect. 



1 These figures are the joint work of Mr. Clark, of the American commission, and Mr. Macoun, of the British 

 commission. They were accepted by Professor Thompson, though made in his absence, as he did not arrive in time to 

 witness the work. After the departure of Mr. Macoun and ourselves from the islands Professor Thompson, assisted by 

 Mr. Lucas, undertook a recount of the live pups. Mr. Lucas's action in the matter was one purely of courtesy, no 

 responsibility for the work of enumeration having been assigned to him by the commissioner in charge. 



In the recount on Kitovi rookery Professor Thompson found 5,534 live pups; Mr. Lucas, 5,577. In a single portion, 

 of the rookery Mr. Lucas found 1,318 pups, whereas Professor Thompson found only 1,217. No effort was made by 

 recounting or otherwise to remove discrepancies. The mean of the two counts, or 5,555, was assumed as the total. To 

 this the dead pups being added, a total of 5,760 pups for this rookery was found as against 5,289 originally counted. 



Afterwards a recount of Zapadui Keef was made. Here, following the same methods, a total of 2,786 pups was 

 found as against the total of 3,041 of the original count. No further recounts were attempted. 



Professor Thompson has seen fit to substitute the results of his recount on Kitovi rookery for the official one, 

 rejecting as unsatisfactory his recount of Zapadni Keef. A comparison of the two counts shows plainly why. Had 

 Professor Thompson substituted both counts, or better yet, had he completed the recount on the remaining: rookeries and 

 used the completed results, his action would have been less open to criticism. 



The results of the recounts are in no sense binding upon the American commission. They bear on their face the 

 evidence of their faulty character, which is strengthened by the admission that one at least is in error. It may he said 

 that Mr. Macoun does not share with his colleague the faith which is placed in the recount. 



We may say that in these recounts no precaution was taken which was omitted in the original count. The 

 conditions of the count by Clark and Maconn were more favorable. The pups were ten days younger, were less active, and 

 were not going into the water. The count was made in a continuous session of five hours, whereas in the recount the 

 rookerv was abandoned for a period at noon, giving room for the possibility of shifting among the pups. 



The grave objection to the recounts, however, rests in the fact that neither Mr. Lucas nor Professor Thompson had 

 had any considerable experience in the work of counting. On the other hand Messrs. Clark and Macouu made their count 

 on Kitovi rookery after having counted 10,000 live pups in 1897 and nearly 25,000 in 1896, to say nothing of 27,000 dead 

 ones and great numbers of cows and harems. 



There is no work in which experience and adaptability count for more than ia the counting of the live pups. The 

 original counts, therefore, stand to the recounts as the work of experts to that of amateurs. It is with great reluctance that 

 we refer to this matter, and we would not mention it were it not that Professor Thompson lias used it to cast discredit on 

 figures undoubtedly accurate and trustworthy. Even with his substitution there is still left by his own accepted figures a 

 positive decline of 9.1 per cent. The difference between this and 12 per cent is of no real importance except that the use of 

 the discrepant figures serves needlessly to weaken the apparent force of evidence drawn from actual enumerations. 



