Here, too, it should be noted that Hart and Spearman mixed a variety of psychoses (even including an imbecile not deteriorated so far as known), both those that do involve actual decay of ability, and those that involve only disturbances of general auxiliary functions, like attitude and effort. Just what would be the effect of this mixing upon the correlations could be told only if we knew how each form of disorder characteristically affects the relationship among mental functions, which is unknown. If mental functions are differently selected for impairment in the different forms of psychosis, then we should expect diminished coefficients of correlation among the insane, because mixing the psychoses would produce inconsistency of rank within the group. If, however, certain functions were deteriorated in all or nearly all of the insane, others remaining intact, or relatively so, this selective enfeeblement would not appear in correlation coefficients. Facts like those observed by Binet and Simon might be obscured by the methods of Hart and Spearman. Moore, working subsequent to Hart and Spearman, limited his investigation to those cases believed by psychiatrists to be characterized by real loss of abilities, the dementias: dementia paralytica, senile dementia, and alcoholic dementia. He tested thirty dements, laborers and tradesmen, and, as controls, six young men from the same occupational group, in the following mental functions: (1) perceiving eight each (in a series) of real objects, pictures of objects, printed words, and spoken words, referring to real objects of ordinary everyday experience; (2) repeating after one exposure of the series as much of it as could be remembered without regard to sequence; (3) after a minute of mental work at calculation, repeating again what could then be remembered of the series. Moore then correlated performance within the group in each of these functions with that in each of the others. The coefficients thus resulting are interpreted as follows: “The average of all correlations of perception with the various memories is .538.... That the average correlation for memory and perception is as high as .538 shows that there must be a common factor present. But its presence does not exclude the existence of special forms of mental ability.” Moore also correlated perceiving with remembering in the functions separately, and remembering immediately with remembering after a minute of distraction. These coefficients are positive, and mostly high, but not perfect.