3. The only valid inference from the phenomena of design would be that of a _phenomenal_ first cause. The inference of a personal Divine Agent or substance from the observation of the mechanism of the universe is invalid. What link connects the traces of mind which are discerned in nature (those _vestigia animi_) with an agent who produced them? There is no such link. And thus the divine personality remains unattested. The same may be said of the divine _unity_. Why should we rest in our inductive inference of one designer from the phenomena of design, when these are so varied and complex? Or grant that in all that we observe a subtle and pervading 'unity' is found, and as a consequence all existing arrangements point to one designer, why may not that Demiurgos have been at some remote period himself designed? And so on _ad infinitum_. But, in the _second_ place, not only is the argument defective (admitting its validity so far as it goes), even partial validity cannot be conceded to it. The phenomena of design not only limit us to a finite designer, not only fail to lead us to the originator of the world, or to a personal first cause, but they confine us within the network of observed designs, and do not warrant faith in a being detached from or independent of these designs, and therefore able to modify them with a boundless reserve of power. These designs only suggest mechanical agency, working in fixed forms, according to prescribed law. In other words, the phenomena of the universe which distantly resemble the operations of man, do not in the least suggest an agent exterior to themselves. We are not intellectually constrained to ascribe the arrangement of means to ends in nature to anything supra-mundane. Such constraint would proceed from our projecting the shadow of ourselves within the realm of nature, and investing _it_ with human characteristics, a procedure for which we have no warrant. Why may not the arrangements of nature be due to a principle of life imminent in nature, the mere endless evolution and development of the world itself? We observe that phenomenon A fits into phenomena B, C, and D, and we therefore infer that A was fitted to its place by an intelligent mind. But suppose that A did not fit into B, C, or D, it might in some way unknown fit into X, Y, or Z,--it would in any case be related to its antecedent and consequent phenomena. But our perception of the fitness or relationship gives us no information beyond the _fact of fitness_. Any other (larger) conclusion is illegitimate.