564 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
however, of the laws of perspective is painfully evident. This speci- 
men is from the lower Magdalenian horizon of the cavern of Gourdan 
(Haute-Garonne). 
In the same layer at Gourdan was found another fragment of rein- 
deer horn with panel engravings that are of more than passing inter- 
est (fig. 14). That the dorsal view presented difficulties perhaps 
even greater than those of the front view is seen in the upper left- 
hand panel. The model in this case 
was a bovidian. This was a daring 
enue artist who sought difficulties he was 
unable to overcome. Neither was he 
afraid to acknowledge failure if such 
it was considered at the time, for his 
signature appears in two places— 
above the left horn and opposite the 
left shoulder. The adjoining panel 
with fish (pike) in profile is also 
signed in two places, but by another 
artist, whose signature, composed of 
an oval pit with four smaller ones 
above it, is not unlike a four-pointed 
coronet. Of the two lower panels 
only the one on the right is adorned. 
The principal figure is that of a small 
antelope running. The body is in 
profile, while the head is turned from 
the beholder. The posterior convex- 
ity at the base of each ear is indi- 
cated, as it was also in the bovidian 
of the upper panel. The head of a 
horse viewed from in front is seen 
just above the antelope. The front 
view of the head alone presents fewer 
Wie. 18 Reindeer viewed from in Gifieulties than that of the entire ame 
front, engraved on reindeer horn. yal, as is attested by engravings on a 
From the lower Magdalenian de- : i 
posits, cavern of Gourdan (Haute. Wand from the middle Magdalenian 
Garonne). After Piette, L’anthr, deposits of the rock-shelter of Mége 
vol. 15, p. 159, 1904. E - 
(Dordogne). The artist’s represen- 
tation of a deer’s head was so successful that it was repeated with 
slight variations four times on the shaft of the slender wand (fig. 
15). Many representations of the front view are so diagrammatic as 
to be scarcely recognizable. Some of the processes that lead to con- 
ventionalism are simply short cuts to the artist’s goal, the goal being 
to convey a given impression. This can often be done better by 
evading difficulties than by meeting them. The paleolithic artist 
