process is ideation; Horwicz and others refer sensation and conation to affection; while other thinkers have tried to reduce the three processes to two. This general disagreement, however, in no way invalidates the criterion of irreducibility. We may be very uncertain whether a given process is itself elemental, or is to referred to some other process. It none the less remains true that until we have good reason to believe that it can be further reduced we must regard it as elemental. Whatever may be the outcome of these various attempts to refer all mental phenomena to a single primary process, the fundamental principle will remain, that whatever shows itself to be irreducible is an element and must be used as a basis for all further investigations.
There is one other point to be noticed with regard to the irreducibility of our ultimates. It is the fact that the work of analysis does not actually cease when they have been discovered. We have said that when we have resolved our mental fact into facts which are themselves irresolvable, our process of analysis is finished. This is true; yet it would not be correct to say that there is no further occasion for analysis. There is need of a second process for the purpose of determining the properties of our elements. If we have agreed, for example, to recognize sensation as ultimate, we may then proceed by means of analysis to discover its various attributes, such as intensity and quality. But this, again, does not shake our faith in the validity of our general criterion of ultimates; for this second analysis is in no sense a continuation of the first process, by means of which we arrived at the elemental sensation. In the first analysis we passed successively from one process to another, finding in each new stage the explanation of the more complex one which preceded it. When we have at length reached a process which we cannot explain by means of any other process, our regress is finished, our element is discovered. Whatever analysis may now be possible, will be entirely distinct from the first and will in no way affect its claim to be complete. The attempt to find an explanation for our process in something else than a process, the effort to go
