THE SOUTHERN SEA LION 127 



Every month of the year except April is represented, so that it is obvious that these 

 skulls could be arranged to form a continuous and progressive series in which no age 

 groups could be distinguished. In order, therefore, to facilitate the identification of such 

 groups by the creation of gaps in the series corresponding to periods of time the speci- 

 mens were divided into "summer" and "winter" groups. 



" Summer" extends from October to March and includes the breeding season (which 

 may also be a time of least growth owing to preoccupation with reproduction), and 

 "winter " extends from April to September. The division is the more easily made owing 

 to the fortunate absence of specimens taken in April and the presence of only two killed 

 in October, both of them being from animals of advanced age. 



Examination of the skulls thus divided has resulted in the distinction of four age 

 groups in addition to those enumerated in 1934, the new groups obviously representing 

 the seventh, eighth, ninth and tenth years of life. The last group must include any 

 animals of more than ten years old, since the skulls in it are in such a condition that it 

 is unlikely that further age groups could be distinguished unless the series were enor- 

 mously increased, and perhaps not even then, since many of the distinguishing cha- 

 racters (e.g. sutures, by fusion) have by this time achieved a condition of uniformity 

 (1934, p. 290). 1 



The characters which have been used to distinguish the groups are those of size, 

 proportion, both of body and skull, and osteological condition. All of these are subject 

 to considerable individual variation. The condition of the teeth, although providing 

 some indication of advancing age, must only be regarded as supplementary evidence, 

 since the degree of wear varies greatly with the habits of the individual under considera- 

 tion (Colyer, 193 1). 



As before, three measurements have been made on each skull: (a) The total length 

 from the most anterior projection of the premaxilla to the furthest projection of one of 

 the occipital condyles, usually the left. If the left were damaged, the right was substi- 

 tuted, (b) The greatest width across the outside of the zygomatic arch, (c) The length 

 from the projection of the premaxilla to the tip of one of the hamular processes. The 

 skull length as a percentage of the body length and the two subsidiary measurements as 

 percentages of the skull length were also calculated: they are displayed in Table II. 



The skull of no. 1161, which was not available for examination before, has been 

 included in this series. It has been remeasured and the length is 260 mm. against the 

 262 mm. previously recorded ; the difference may be due to the use of a different 

 instrument, but since it is nearly seven years since the skull was collected it seems pos- 

 sible that it may have shrunk through continuous drying. The difference of 2 mm. is 

 only 077 per cent and cannot be regarded as of great importance. 



In comparison with the skull of the immature, that of the adult shows an increase 

 in the relative size of the facial region and an alteration of the palate to the elongated 

 and arched form which is accompanied by a downward growth of the ventral plates of 



1 In order to avoid too lavish recurrence of my name in this paper, references to my previous report are 

 inserted in this fashion. 



