Now let us exert energy and disarrange a group. Those things that were or might have been tangible in the former position, are no longer so, but they may still be seen, heard, or even smelt. The bright colours have however somewhat faded, the size has shrunk, some of the details are lost. Here is a lapse from perspective completeness. It is indicated, not as in the first two categories by mechanical cutting away of mass and circumstance, but by deterioration all over the object. We seem to be thrown out of focus in relation to it, and the perspective degradation may increase until the object has dwindled to a speck and finally disappears altogether.

The judgment in this category consists in observing the kind and degree of degradation to which things are liable in perspection. In addition to change in size, brightness, detail and loudness, which have been already mentioned, occultation as in the second category can be used as an indirect datum. An object which eclipses another is invariably more perfect perspectively than the object eclipsed. The motion of objects has also to be taken into account. As objects degrade their movements slacken, and recover power as the objects are restored.

By attending to all these indications and checking each by the rest, we have the elements of a fairly accurate inference as to comparative perspective condition. We have constant practice in this sort of thought with frequent opportunities of verifying our conclusions; penalties are annexed to failure and rewards to success. It is no wonder then that in the course of years we become expert in judging of perspective condition, so that when confronted with a natural group we can estimate almost instantly the degree in which each object falls short of perspective integrity.

The result of this practice is that on perceiving a natural group of many objects, we graduate them according to the perspective deterioration which each exhibits, and for greater precision we figure the perspective difference as an interval between the objects—an imaginary interval modelled on the true interval of association. The object on a distant horizon is visually as near as the ground we can touch by stooping, but in this imaginary group the former is placed at the far end of the line and the latter at the near end, and between them are ranged the other objects each at a point corresponding to what we suppose to be its perspective distance. That is how a landscape acquires depth. Space outwards is an ideal imitation of real lateral interval. It is the measure and expression of perspective defacement.

From what has been said it follows that the near objects will be relatively large, clear, and lively in motion, while the far will be small, dull, and slow, but this rule is liable to many exceptions which can only be learnt by experience.

On the analogy of the other forms of inference—which consist in completing imperfect things by reference to others more perfect—the essence of an argument in perspection is the power to imagine an object which is perspectively defective, brought up to the tactual range and displaying all the qualities it would possess in that position. This is done by comparing it with the idea of the same or a similar object experienced at the tactual range; and is done for an ulterior purpose, like all other intellectual operations. A great part of our material happiness consists in the exercise of the short senses (taste, touch and smell), and the chief use of perspective reasoning is to enable us to judge of the energy required to bring a distant object near for close perception. We have therefore to observe our energic fluctuations in conjunction with perspective change, if we would extract the utmost practical benefit from this category. The perspective inferences are none the less useful after we discover that they are not intuitions, and that the completeness we imaginatively assign to distant objects has no existence until we exert the corresponding energy.

A landscape being rendered perspective we can determine the perspective state of any new object that may enter it, by reference to the objects adjoining it, and this though the object be of a species quite unknown to us and which therefore, by itself, would afford no clue to its perspective distance.

The imaginary interval we place between objects of different perspective effacement, can be expressed in terms of exact lateral measurement. This is done by developing and measuring the associative groups represented in the perspective group. Supposing we wish to get an exact definition of the perspective condition of a mountain relative to a certain station, we can, from that station, develop all the natural groups up to the mountain (walk over the ground) and measure the lateral intervals and masses disclosed. The total measurements will be a definition of the mountain's perspective distance in terms of true associative distance. That is what we mean by saying a mountain is ten miles off. It is not really ten miles off—it is not an inch off. But to render it tactually perfect we should have to expend an amount of energy equal to 17,600 times the energy required to move from one associative object to another a yard apart from it laterally. If we practise the mileage scale in conjunction with the perspective indications, we may acquire the art of expressing in miles, though not measured, the distance of objects estimated from purely perspective data, but few can do this with any near approach to exactness[14].

The realistic three-dimensioned space is a combination of the true interval of association and the false interval of perspection. This generates an idea resembling the capacity or vacancy in a room or vessel, and thus it is supposed that objects occupy a sort of universal room without walls, floor, or ceiling. It is however the enclosing objects which make a room, and when they are abstracted there remains nothing. The universal room is therefore nothing—a myth. It is a useful working theory for common purposes, but in philosophy it is superfluous and obstructive.