On the other hand, he w i l l aggregate without necessarily employing such prejudices and standards. Of course, this is not a wholly clear distinction. I t is likely that males w i l l more readily bond wi th persons of appropriate ascribed status - just as they w i l l tend to mate particular females who have certain ascribed characteristics. But, broadly speaking, aggregation involves no process of choice. A l l the males of the right age in a community, be they schoolboys or bison, aggregate inevitably in simple spatial terms and possibly in ways that are more complex. 1 Furthermore, aggregation differs from 'mobbing behaviour', 2 which does not appear to have a unisexual basis but which otherwise clearly involves social processes similar to aggrega­ tion. As Lorenz writes:

Social animals . . . take every possible chance to attack the 'eating enemy' that threatens their safety. This process is called 'mobbing'. Crows or other birds 'mob' a cat or any other nocturnal predator, i f they catch sight of i t by day. 3

In Lorenz' s view, very close herding and intra-specific aggression tend not to occur in the same species. His conception of the relationship between bonding behaviour and aggression w i l l be discussed below.