54 



NORTH AMERICAN FAUNA 74 



flippers but were present in the hairless "sensory" areas of the oral opening. In 

 these sensory areas, a thick, almost granular exudate was excreted from the 

 slender clusters of sebaceous glands; the sweat glands there were greatly reduced 

 and apparently not functional. 



In the haired skin of walrus calves and adults, the ducts of the sweat and 

 sebaceous glands generally open into the hair canals at the same, rather than at 

 different levels. Conversely, in the body skin of the Otariidae, the nearest rela- 

 tives of the walruses, the orifice of the sweat duct is by far the more distal; in 

 most of the Phocidae, the sweat orifice tends to be the more proximal (Ling 

 1965). The significance of this difference between the otariids and phocids, in 

 relation to phylogeny and function was discussed recently by Ling (1965) and 

 Mitchell (1967). Both investigators recognized that there are interspecific and 

 topographic variations in the positions of the ducts in both groups, but concluded 

 that the more proximal position in most Phocidae probably is functionally 

 related to the less dense pelage. On that basis, the walrus might have been 

 predicted to have sweat orifices more proximally situated than those of any other 

 pinniped, since it is the most sparsely haired of all (Scheffer 1964). My finding 

 that the sweat orifice is consistently distal to those of the sebaceous ducts in fetal 

 walruses implies phyletic relationship to the Otariidae (recently confirmed by 

 Tedford 1976, and Repenning and Tedford 1977). The shift in relative positions 

 of the ductal orifices before birth implies pre-adaptation to the sparsely haired 

 condition in later life. One can infer from this that ancestral walruses were more 

 densely haired than those living today, and that the trend toward sparser pelage 

 is a relatively recent development, convergent on the Phocidae. Because the shift 

 in position of the ductal orifices takes place rather late in fetal development, it 

 may be a recent adaptation that is still under way. 



Pelages and Molts 



Prenatal Pelage and Molt 



In four fetuses with crown-rump lengths of 10.6, 13.2, 17.8, and 18.6 cm, 

 taken in August, minute primordia of pelage hair follicles were present in the 

 skin on the torso, neck, and head; those on the cheeks, snout, and lips were 

 further advanced in development than any others. In the next larger fetus, 

 20.8 cm in crown-rump length and taken in September, elongate hair canals 

 and primordial sebaceous glands were present. In a 28.5-cm fetus, taken in early 

 October, about 3.5 months post-implantation, hair cones of the first (lanugo) 

 pelage were developing. 



Seven fetuses taken near the end of November and in the first half of December 

 (5.5 to 6 months post-implantation) had a dense coat of fine whitish to silvery 

 hair and well-developed mystacial and superciliary vibrissae (Fig. 36). 

 Crown-rump lengths of these specimens ranged from 45 to 63 cm. The pelage 

 hairs, 3 to 6 mm long, were very slender, sinuous, and circular to elliptical in 

 cross section. They contained no medulla and only traces of pigment, usually in 

 the distal third of the shaft. Generally, a single hair was present in each hair 

 canal. When there were more than one per canal, the hairs arose from branched 

 follicles of about equal length and were virtually identical, hence not classifiable 



