Polynesian Dog — Luomala 
207 
few in New Holland. These dogs, he adds, had 
short legs, long backs, pricked ears, an occasional 
howl but no bark. They were shy towards 
strangers and stupid. They were eaten. Walther 
drops the Low Islands (the Tuamotus are prob- 
ably meant) from the list and substitutes the 
Sandwich (Hawaiian) Islands. The references 
to the Sandwich Islands and to the dogs being 
long-backed and short-legged (he does not in- 
clude the fact that the legs were often crooked) 
indicate that Walther, although he does not give 
his source, was probably familiar with King’s 
description. He probably includes New Holland 
with the islands having his meridionalis because 
he interprets Pennant’s ambiguous introductory 
statement to refer only to the first of the 
varieties described. 
Walther has another Polynesian subvariety 
which he calls Cams familiaris villaticus, novae 
Zeelandiae, the New Zealand dog. He describes 
it from sources that he gives as Bechstein (1799, 
I: 258) and George Forster (1778, I: 165). 
This dog, he says, resembled a shepherd’s dog, 
and had long silky hair, pricked ears, and dif- 
ferent colors. It was found in New Zealand 
where it was fed almost exclusively on fish and 
kept only for its hide, from which festive attire 
was made. It was not eaten. It was very stupid, 
had a poor sense of smell, and rarely barked but 
only howled now and then. 
Two more South Pacific dogs classified by 
Walther (1817: 21-24) are the Australian 
dingo, Cams familiaris villaticus, novae Hol- 
landiae , and the "fox-like dog of New Guinea,” 
Canis familiaris villaticus, novae Guineae. He 
gives as sources Bechstein (1799, I: 260) and 
Forrest’s ("Forster”) German edition (1782: 
121, 126). An error in his book which regularly 
gives Forster when Forrest is meant is con- 
tinued by later writers. 
Walther obviously then makes explicit and 
gives binomial classifications to four Pacific vari- 
eties that he distinguishes and that are not so 
explicit in Pennant’s account. Moreover, he 
(1817: 21-24) puts all four in the same larger 
classification villaticus as four Old World breeds, 
to make a total of eight "national breeds” in 
C. familiaris domesticus. The Old World breeds 
in villaticus are the common, black, long-haired 
German house dog with curled-over tail, ger- 
manicus (one of Walther ’s lists inserts the Lap- 
land dog next ) ; the Kalmuk dog, calmuccorum; 
the Pyrrenes dog, pyrenaicus; and that of Pom- 
erania, pomeranus. Walther leaves no doubt that 
he regards the Pacific dogs as belonging to the 
same genus, species, and variety ( Canis familiaris 
villaticus ) as the Old World breeds of domes- 
ticated house dogs. So far as I can determine, 
he is the first even to attempt to classify the 
Polynesian dog by genus and species. 
H. G. L. Reichenbach (1836: 46, pi. 72) 
does not describe the dog he classifies as Canis 
otahitensis, but his sketch (Fig. 18) shows some 
dogs whose most remarkable feature is a long, 
flowing tail like that of a horse. The source of 
the artist’s inspiration is not stated. Perhaps 
Reichenbach describes the dog and cites his 
authority on page 141 in his "Naturg. Raubth.” 
(perhaps, in full, Naturgeschichte Raubsdu- 
gethiere ) which I have not seen but which Leo- 
pold J. Fitzinger ( 1867 : 400-401 ) quotes with- 
out giving the complete title. Fitzinger states 
that Reichenbach gives two other classifications 
of the Polynesian native dog. They are Canis 
familiaris orthodotus and Canis familiaris ortho- 
tus otahitensis . The spelling of the sub-species 
varies. Reichenbach (1836: 22) classifies the 
dingo both as Canis Dingo or, according to Fitz- 
inger (1867: 817), Canis familiaris orthotus 
Dingo ("Naturg. Raubth.”: 147, 354), and as 
Chrysaeus Australiae ("Naturg. Raubth.”: 366). 
Consideration of Fitzinger’s own classifica- 
tions will be deferred until later. 
FIG. 12. Detail of Figure 11 to show a long-legged, 
flop-eared, spotted dog with a long muzzle and short, 
half-curved tail. 
