“There are,” said John, “two kinds of creatures that I am well acquainted with—­the one a land animal, the other a water one—­the red-deer and the salmon.  In October the deer ruts, and the salmon spawns.  The deer begins soonest, high up among the hills, particularly in frosty weather; so does the salmon begin to spawn earlier in frosty weather than in soft.  The master hart would keep all the other harts from the hind, if he could; and the male salmon would keep all the other males from the female, if he was able.”—­P. 60.

We do not think, however, that Mr Scrope’s comparative reference to the upper and lower portions of the Tay affords a satisfactory or conclusive test.  The higher parts of almost all rivers (including, their tributaries) constitute the favourite spawning places, from other causes than “by reason of the cold;” and the question should be tried, not by comparing two different districts of the same river, but all the portions of one river, with the entire course of another of dissimilar character.  The exceptive clause in Mr Loch’s proposed act in flavour of the river Ness, certainly stood upon the supposition of that river being an early one for the breeding salmon, as well as the new-run winter fish; for it enacts not only that the Ness should open more than a month earlier than its neighbours, but also that it shall close more than a month before them.  This latter restriction would of course be useless and impolitic, if the parent fish were not conceived to be about to spawn.  But it should also be borne in mind, that the same causes (such as the extent and depth of feeding lakes) which produce a higher temperature in winter, cause a lower one in summer and the earlier part of autumn, and that shallow upland streams are warmer during the latter periods than those which flow from deeper and more affluent sources.  We believe that the fish of all rivers spawn soonest on the higher portions of their water courses, whether these be comparatively warm or cold.  The earliest individuals are in general such as have escaped the nets and other accidents below, and have made their watery way in good time to proper spawning places.  In several rivers with which we are acquainted, a great majority of the breeding fish ascend in August and September.  But many of those which make their appearance in July, would be early spawners if they were allowed to escape the various dangers which beset their path in life—­almost all the salmon of that month being captured by one means or another.  Mr Young, in our MS. notes already quoted, states, in regard to the range of the breeding season, that he has seen salmon perfectly full of spawn, ascending the rivers in October, November, December, January, and February.  Now the fish of the last-named month may have spawned as late as March, although our correspondent adds that he has never seen fish on the spawning beds later than February, nor earlier than September.  He has seen them in the act of spawning in these and all the intermediate months.