incidentally to hunt and fish.  But in the fall, when a new ice bridge across the Strait releases the reindeer from their enclosed pasture on the island, the Samoyedes withdraw southward, and the merchants with their wares to Archangel and other points.  This has gone on for centuries.[328] On the Briochov Islands at the head of the Yenisei estuary Nordenskiold found a small group of houses which formed a summer fishing post in 1875, but which was deserted by the end of August.[329]

[Sidenote:  Altitude boundary zones.]

An altitude of about five thousand feet marks the limit of village life in the Alps; but during the three warm months of the year, the summer pastures at eight thousand feet or more are alive with herds and their keepers.  The boundary line of human life moves up the mountains in the wake of spring and later hurries down again before the advance of winter.  The Himalayan and Karakorum ranges show whole villages of temporary occupation, like the summer trading town of Gartok at 15,000 feet on the caravan route from Leh to Lhassa, or Shahidula (3,285 meters or 10,925 feet) on the road between Leh and Yarkand;[330] but the boundary of permanent habitation lies several thousand feet below.  Comparable to these are the big hotels that serve summer stage-coach travel over the Alps and Rockies, but which are deserted when the first snow closes the passes.  Here a zone of altitude, as in the polar regions a zone of latitude, marks the limits of the habitable area.

[Sidenote:  “Wallace’s Line” a typical boundary zone.]

The distribution of animals and races shows the limit of their movements or expansion.  Any boundary defining the limits of such movements can not from its nature be fixed, and hence can not be a line.  It is always a zone.  Yet “Wallace’s Line,” dividing the Oriental from the Australian zoological realm, and running through Macassar Strait southward between Bali and Lombok, is a generally accepted dictum.  The details of Wallace’s investigation, however, reveal the fact that this boundary is not a line, but a zone of considerable and variable width, enclosing the line on either side with a marginal belt of mixed character.  Though Celebes, lying to the east of Macassar Strait, is included in the Australian realm, it has lost so large a proportion of Australian types of animals, and contains so many Oriental types from the west, that Wallace finds it almost impossible to decide on which side of the line it belongs.[331] The Oriental admixture extends yet farther east over the Moluccas and Timor.  Birds of Javan or Oriental origin, to the extent of thirty genera, have spread eastward well across Wallace’s Line; some of these stop short at Flores, and some reach even to Timor,[332] while Australian cockatoos, in turn, have been seen on the west coast of Bali but not in Java, Heilprin avoids the unscientific term line, because he finds his zoological realms divided by “transition regions,” which are intermediate in animal types as they are in geographical location.[333] Wallace notes a similar “debatable land” in the Rajputana Desert east of the Indus, which is the border district between the Oriental and Ethiopian realms.[334]