But even if the foremost importance of emotional elements and the possibility of treating them as collective feelings were granted, there is another objection to be met. Granted that these elements are actually quite essential in determining family relations, it might be objected that they are too shapeless and indeterminate in themselves to be of any practical use in scientific research, especially if our theories have to be based upon ethnographic observations in which the more tangible and the more unambiguous the facts chosen, the less the risk of being misled. Now, are not feelings of the most indeterminate character, the most misleading, and the most difficult to ascertain? In fact, the theory of feelings and emotions seems to be the least developed in individual as well as in social psychology. Especially it might be suggested that to pursue the investigation on double lines is useless; feelings always find adequate expression in ideas, in fact crystallize in them.[557] Without trying to give a general answer to these objections, they may be met as regards the special case under discussion. In Australia, as a matter of fact, they do not hold good. For our knowledge of the sentimental side of parental kinship is much better and much more determinate than our knowledge of any other aspect of this relation. [557] It seems needless to add that the deep connection and mutual dependence of both feelings and ideas is perfectly acknowledged. This is not the place, of course, to pursue any detailed psychological investigations. I would like to remind the reader that all that is said here must be judged by its application to the Australian facts given below. In higher societies where art, poetry and thought lend themselves much more to the expression of feelings, the former afford objective documents of the latter. In low societies we must look for such objective documents elsewhere, in different sets of facts.