museum and the others raced off among the upper branches of the trees.  Then we came on a party of coatis, which look like reddish, long-snouted, long-tailed, lanky raccoons.  They were in the top of a big tree.  One, when shot at and missed, bounced down to the ground, and ran off through the bushes; Kermit ran after it and secured it.  He came back, to find us peering hopelessly up into the tree top, trying to place where the other coatis were.  Kermit solved the difficulty by going up along some huge twisted lianas for forty or fifty feet and exploring the upper branches; whereupon down came three other coatis through the branches, one being caught by the dogs and the other two escaping.  Coatis fight savagely with both teeth and claws.  Miller told us that he once saw one of them kill a dog.  They feed on all small mammals, birds, and reptiles, and even on some large ones; they kill iguanas; Cherrie saw a rattling chase through the trees, a coati following an iguana at full speed.  We heard the rush of a couple of tapirs, as they broke away in the jungle in front of the dogs and headed, according to their custom, for the river; but we never saw them.  One of the party shot a bush deer—­a very pretty, graceful creature, smaller than our whitetail deer, but kin to it and doubtless the southernmost representative of the whitetail group.

The whitetail deer—­using the word to designate a group of deer which can neither be called a subgenus with many species, nor a widely spread species diverging into many varieties—­is the only North American species which has spread down into and has outlying representatives in South America.  It has been contended that the species has spread from South America northward.  I do not think so; and the specimen thus obtained furnished a probable refutation of the theory.  It was a buck, and had just shed its small antlers.  The antlers are, therefore, shed at the same time as in the north, and it appears that they are grown at the same time as in the north.  Yet this variety now dwells in the tropics south of the equator, where the spring, and the breeding season for most birds, comes at the time of the northern fall in September, October, and November.  That the deer is an intrusive immigrant, and that it has not yet been in South America long enough to change its mating season in accordance with the climate, as the birds—­geologically doubtless very old residents—­have changed their breeding season, is rendered probable by the fact that it conforms so exactly in the time of its antler growth to the universal rule which obtains in the great arctogeal realm, where deer of many species abound and where the fossil forms show that they have long existed.  The marsh-deer, which has diverged much further from the northern type than this bush deer (its horns show a likeness to those of a blacktail), often keeps its antlers until June or July, although it begins to grow them again in August; however, too much stress must not be laid on this fact, inasmuch as the wapiti and the cow caribou both keep their antlers until spring.  The specialization of the marsh-deer, by the way, is further shown in its hoofs, which, thanks to its semi-aquatic mode of life, have grown long, like those of such African swamp antelopes as the lechwe and situtunga.