considerations; and to decide the species I gather up what help I can from the hints furnished by the size and shape of the tree, the time of coming into leaf, the colour of the flowers; or perhaps, if I am lucky, a bit of leaf or blossom may be blown within reach of me. But so great is our triumph, and so complete sometimes our success in finding elaborate substitutes for a simple walk or change of place, that we are very apt to overlook how almost ridiculously simple the end we aim at would often be, if only our faculties of locomotion were less restricted than they unfortunately are at present. There are myriads of facts about which if any doubt is felt it is dispelled at once by some one just going to look at the things; there are myriads of other facts, in all essential points often just as simple, which because unfortunately we cannot go and look at them,' task the highest powers of thought of our greatest philosophers, and the most exquisite skill of our instrument makers. If we want to know how hot it is at the Antipodes we go there ourselves, or send some one else there, with a thermometer. But if we want to know how hot it is half way there, viz., at the centre of the Earth, or indeed whether it be very hot at all there, we are led into the most intricate questions of physics and mathematics, through which at present we can find no certain way.

Similar considerations apply also to the case of time, and, if possible, even more strongly; for here, for one reason or another, we seem to be more prominently occupied with the mere description of events (which is equivalent to their direct observation) and less on the whole with their analysis and generalisation. Very many of the facts which the ordinary historian toilsomely works out by elaborate comparison of records, and inference in filling up the gaps which they leave between them, are such as he could settle almost at once and completely to his satisfaction, if only he could just step back into the time in question. It may be replied that though he cannot go to the events the events can come to him, through the testimony of witnesses and other records. It is clear however that this one-sided process is a very poor substitute. The witness is not of our own selection: perhaps he does not know what are the important points which he ought to observe; and, one may almost say, he knows that he cannot be cross-examined, and is therefore subject to hardly any check. A few hours spent in personal observation upon the spot by a critical historian himself, would sometimes be worth a whole volume compiled by contemporary witnesses. This must especially be the case where we are concerned with general dispositions and tendencies rather than specific facts. People have disputed, and will continue to dispute, for instance, whether and to what extent our age is more moral than former

ages. What a light we should gain upon this point if only some London police magistrate, some doctor in general practice, or some shrewd man about town, could go, with proper introductions, into some other century, and live for a few months on easy terms with its inhabitants!

But many of the past facts which we want to decide can rest upon no personal testimony. We cannot appeal to the witness, because he was never there. And yet the facts themselves may be of just the same kind as those every-day phenomena in the decision and estimate of which any ordinary person is nearly as good a judge as any other. Such inquiries as whether the earth was once fluid, whether the glacial period prevailed more than once, and how far southwards it extended, are not in themselves more difficult of decision than to decide whether the lava from some volcano is fluid, or whether there are two winters annually in the Arctic regions. So far as any difficulties in the phenomena themselves are concerned, apart from our means of getting to know them where we are now, these questions could be settled at once by any witness as good as an ordinary skipper, without the least hesitation or doubt. He would merely have to tell us what he had seen and felt, and the matter would be set at rest at once.

The foregoing considerations are obvious enough when pointed out; and they would probably be perfectly familiar to every one, were it not that, as already remarked, the excessive ingenuity and complication of the various substitutes which have been discovered for the unattainable visit of observation tend to conceal from view the extreme simplicity of that end in itself. In saying this it is not intended to imply that the processes of inference at all resemble observation. On the contrary many of them are of a highly abstract and generalised kind, and they are mostly carried on more or less symbolically. But they ground or result in an imaginary observation. This is also the form which their concrete applications take, and we may fairly say that we have not attained proper grasp of them until we can mentally reproduce them in this way. We must be able to individualise or picture the results to ourselves before we can be said properly to know them, and to do this is clearly to take an imaginary observation of them.

We may see then already how important would be a complete control over space and time, as for want of a better form of expression we have ventured to call it, for all purposes which concern our inferences. Any such control, if really complete, would tell at once by superseding all need for observation; for why carry on in our study a painfully elaborate and circuitous process when the direct process for which it was

meant to be a substitute was itself within our power? And indirectly it also tells in the same way by both strengthening and simplifying the processes of reasoning. The wider the basis of observation from which we start the better grounded, as a rule, are our conclusions, and the shorter the processes of getting at them. Every bit of extra power therefore that we could gain over these two all-pervading conditions of things would diminish the sphere and lighten the work of inference. Before looking closer at the details, let us just ask in a word or two what it is that we wish for; in other words, what are the requirements for that power of observation which if complete would render all inference superfluous, and which in proportion as it approximates towards completeness so powerfully aids our inferences? These requirements seem reducible to the two following-regard being had to the nature of our faculties and the general conditions under which we have to employ them: power to move about as freely as we may wish in space or time, and power to enlarge space and time to any extent we may need. The sense in which this latter requirement has to be understood will be more fully discussed in the sequel.

Let us begin with the former, viz., our power of locomotion (the reader will observe that we are obliged to use, in many cases, space-words for time-ideas, and vice versa, from inadequacy in ordinary terminology). What our powers are in this respect as regards space, every one knows. Within very small limits we can move ourselves, or the objects with which we are concerned, up and down and about, in three dimensions, as we please. Within wider limits, viz., that of the surface of the globe, we are restricted to two dimensions. Beyond that again we are hampered still further by being confined to one dimension only, our motion along that even being quite beyond our own control. This of course refers to the motion of the earth round the sun, or any further motion that our system may have through space. Even this however, as an aid to knowledge, counts for something, so that we should not receive it without gratitude; for some of our knowledge of the shape and magnitude of the visible heavens depends more or less upon this power of linear movement.

But though every one knows the nature and limit of our powers in this respect, it is only those who have given some attention to psychology who can at all realise their importance in the processes of gaining knowledge, or the extent therefore to which our powers of inference are crippled by their very partial and one-sided development. Let us then take an example and look at it a little more in detail; and instead of beginning with some broad and concrete circumstance or aggre

gate of circumstances, it will be better to commence with a minute one. I am inspecting some small object, say a penknife of unfamiliar construction, and I want clearly to understand its mechanism, size and shape. How is it that I am able to do this so completely and accurately? Mainly on this ground, that I am able to turn it about at will so as to present any face towards me, and to put it at any required distance far or near. With this, however, must be combined an important consequence of this power of adjustment, viz., the power of looking at the same point of the object again and again as often as we please. This consideration is a very important one. My actual range of observation, as every one knows, is at any one moment extremely minute, almost indefinitely so, the merest point only being presented to the eye. But by running the eye repeatedly over the main outlines, by frequent recurrence to points already once looked at so as to bring them into connexion with the remaining points, we succeed in building the various parts up into one connected whole. We then consider that we have understood or taken the whole object in. To attain to this end it clearly does not much matter whether the power of local movement is on our side or that of the object, whether we turn about it or make it turn all sides to us. Motion being merely relative, either of these amounts to that control of space of which we are talking; and it is generally more convenient, when we can do so, to move the object rather than to move ourselves, for much the same reasons as make the turner prefer that the object should rotate under his tool rather than he be at the trouble of moving his tool round the object. But without this power of freely moving the thing relatively to ourselves or ourselves relatively to the thing, we should have extremely slight opportunities of getting familiar with the mutual arrangement of the various parts of any object small or great.

Now this state of powerlessness represents almost exactly our relation to events in respect of time. We are bound, as we all know, to go steadily forwards: we have no power to stand still, go sideways or backwards. It is easy to perceive how serious a hindrance is thus caused in our investigations. Suppose we are examining some small time-event. If it present itself in the form of some process which is entirely within human control we may probably be able to stop it altogether at some arbitrary point, or to invert its order of occurrence. This comes to very much the same thing as being able to turn the pen-knife over in our hands-in point of fact we are making use of our superior powers of space-locomotion and are substituting movement here for movement in time. The comparison therefore is not quite a parallel one; for it is not really the same event which we thus

turn about temporarily, but only an exactly similar one. It is as if we had no power to move our knife, but by looking about the room could observe any quantity of other knives differently disposed close by, but all exactly similar in their construction. We should thus contrive to supplement our impressions gained from one of them by those gained from the others. When however the operation under observation is a natural one, and therefore outside of our direct control, we are in general quite powerless to do anything of this kind. It is a very difficult thing to find what would popularly be called the same event' twice overthat is, two distinct events alike in all essential respects, but differing from one another by being each of them in exactly the desired stage of development. Different stages of development can often be secured readily enough, but they labour under the essential defect of discontinuity: that is, we cannot secure any desired exact intermediate stage at will: to say nothing of the obvious difficulty of securing that they shall really be alike in all other essential particulars. Suppose, for instance, we are examining the process of germination of a seed. We may find it the best plan to grow a great many of them, and then select some of them for examination, thus securing that there shall be some of them in almost all the successive stages of development. But this seems to fall just as far short of what we really want as would an offer to look at a variety of knives, very much like one another, lying in various directions about the room, but without permission to stir from our position or touch any one of them, fall short of the advantage of handling any one of them at leisure, and turning it about into any desired position. What a gain, for instance, it would be to any student, say of embryology, if he could put his object under a microscope and then shift it a few minutes or hours forwards or backwards in time according to his choice! Forwards of course he can go, or rather must go, as things are now, provided the processes in question are not brought to a stand-still by means of the observation itself. But apart from the tedious uniformity of the pace with which that progress may now be carried on, there is the fatal defect that we cannot pause at the critical stages, and recall and re-observe them at our leisure. What we want is the power to stop still and to go backwards whenever we please. Compare the position of a man who has got an intricate argument in writing before him, with that of one who can merely listen to it as it is repeated to him in order, and we shall realise the difference between what our powers of observation now are and what they might have been had these things been other than they are. What we want in fact is a microscope with a double set of stage-screws; one set to move the stage about as is now done,