almost entirely filled with the works of Reynolds and Gainsborough, while other examples of the same painters will be found in the welllighted vestibules which bound the central staircase.

The west wing of the Gallery, including the rooms numbered from XVIII. to XXII., is entirely given up to the remaining portion of the British School, beginning with the works of Copley, Crome, Blake, Wright of Derby, Constable, and their contemporaries down to pictures of comparatively modern date, arranged as nearly as possible in chronological order, though the largest of Turner's landscapes were necessarily reserved for the last room, which from its size is the only one in the west wing capable of containing them.

Those English cabinet pictures which, for want of space, had been temporarily exhibited in a ground-floor room, have been brought upstairs, and now take their proper place in the national collection.

The total number of rooms containing oil paintings at the National Gallery is now twenty-two. Of these, five are new and had to be filled with pictures, while thirteen of the others had to be either entirely or partially rehung. The work was begun early in May and was completed before the end of June. During that period the Gallery was kept open daily to the public, with the exception of the particular room or rooms which happened for the time to be dismantled.

Those who have had some experience of the difficulties which attend the arrangement of a private collection will best appreciate the care and attention required in executing the same task for a public gallery containing about 1,200 works, many being of exceptionally large size.

It is generally admitted that a symmetrical disposition of pictures, when practicable, is more agreeable to the eye than when they are hung in a haphazard fashion, and if mathematical uniformity of grouping were the only object in view, that object, in dealing with a miscellaneous collection, might easily be secured. But whenever pictures are classified under Schools,' the possibility of ranging them according to size is at once greatly restricted. Nor is this all. When two or more of similar dimensions have been selected for a group, it not unfrequently happens that the nature of the subjects treated, the chromatic quality of the works, or the style of their execution, is such as to render their juxtaposition incongruous. brilliant and gaily coloured picture hung near one which is painted in a sober key will sometimes take all life out of the latter, and reduce it to the level of a work in monochrome. On the other hand, to group together a series of sombre-toned pictures is apt to produce a gloomy effect, and the hanger has therefore to steer between two extremes. Again, for the purpose of study, it is no doubt convenient that the works of a particular master should hang side by side. But the varying scale of such works, their dissimilarity of shape-horiVOL. XXII.-No. 130. 3 K

A

zontal, upright, square, circular, oblong, or arch-headed, for instance --will often present obstacles to such a plan.

And even where conditions of size, colour, and subject are favourable to the close association of certain pictures, it may happen that while some of them deserve the best place that can be found on the walls, one at least of the group is of inferior quality, and cannot be allowed to occupy space which may justly be claimed for a work of higher merit though less consonant in style.

Last, though not least, the requirements of students have to be borne in mind. The pictures which they select to copy may not be always in the first rank of art, but they are generally popular works, and when the object in view is to produce a saleable copy, or perhaps to execute a commission, some pardonable dissatisfaction would ensue if the original were to be hung in an inconvenient place. In the British School, therefore, a careful adjustment was necessary to prevent the chance of a grievance on that score.

When all these various demands have been met, or at least considered, there remains the most formidable problem of all-how to satisfy that inevitable variance of taste which characterises a critical public. Enthusiastic admirers of a special School or particular Master are apt to disregard the importance of works which represent other phases in the history of pictorial art, and probably no two persons would agree as to the relative merits of even a dozen pictures.

To attempt the arrangement of a large collection in such a manner as to gain the approval of every individual connoisseur would indeed be to undertake a hopeless task. In this and similar cases, all that can be deemed possible is to accomplish the work with such due regard for consistency of purpose as will disarm criticism from those who, recognising the aim in view, and aware of the impediments to its fulfilment, are best qualified to judge of the result.

CHARLES L. EASTLAKE.

THE TIME IT TAKES TO THINK.

ALL science is partly descriptive and partly theoretical. Care must, however, be taken lest too much theory be built up without sufficient foundation of fact, or there is danger of erecting pseudosciences, such as astrology and alchemy. The theories of the conservation of energy and of the evolution of species are more interesting to us than the separate facts of physics and biology, but facts should be gathered before theories are made. The way of truth is a long way, and short cuts are apt to waste more time than they save. Psychology is the last of the sciences, and its present business seems to be the investigation of the facts of consciousness by means of observation and experiment. Everywhere in science experiment is worth more than observation; it is said that the evidence in pathology is so contradictory, that almost anything can be proved by clinical cases. Psychology, owing to its very nature, must always depend largely on observation for its facts, and some progress has been made in spite of the difficulties lying in the way of introspection and the correct interpretation of the actions of others. The application of experimental methods to the study of mind is, however, an important step in advance, and would seem to be a conclusive answer to those who, with Kant, hold that psychology can never become an exact science. I propose explaining here how we can measure the time it takes to think, and hope this example may show that the firstfruits of experimental psychology are not altogether insignificant or uninteresting. Just as the astronomer measures the distance to the stars and the chemist finds atomic weights, so the psychologist can determine the time taken up by our mental processes. It seems to me the psychical facts are not less important than the physical; for it must be borne in mind that the faster we think, the more we live in the same number of years.1

It is not possible directly to measure the time taken up by mental processes, for we cannot record the moment either of their

The results I am about to give are based on experiments, detailed accounts of which I have printed in recent volumes of Mind, Brain, and Philosophische Studien.

beginning or of their end. We must determine the interval between the production of some external change which excites mental processes, and a movement made after these processes have taken place. Thus, if people join hands in a circle, and one of them, A, presses the hand of his neighbour B, and he as soon as possible afterwards the hand of C, and so on round and round, the second pressure will be felt by each of the persons at an interval after the first, the time depending on the number of people in the circle. After the hand of one of the persons has been pressed an interval very nearly constant in length passes before he can press the hand of his neighbour. This interval, which we may call the reaction-time, is made up of a number of factors. A period elapses before the pressure is changed into a nervous message or impulse. This time is very short in the case of touch; but light working on the retina seems to effect chemical changes in it, and these take up some little time, probably about sec. After a nervous impulse has been generated it moves along the nerve and spinal cord to the brain, not travelling with immense rapidity like light, but at the rate of an express train. In the brain it must move on to a centre having to do with sensation, where changes are brought about, through which a further impulse is sent on to a centre having to do with motion, and a motor impulse having been prepared there is sent down to the hand. Another pause, to To sec., now occurs, while the muscle is being excited, after which the fingers are contracted and the reaction is complete. The entire time required is usually from to sec. The reactiontime varies in length with different individuals and for the several senses, but as long as the conditions remain the same the times are very constant, only varying a few thousandths of a second from each other. One may wonder how it is possible to measure such short times and with such great accuracy. It would not be easy if we had not the aid of electricity; but when it is called to mind that a movement made in London is almost instantaneously registered in Edinburgh, it will not seem inconceivable that we can record to the thousandth of a second the instant a sense-stimulus is produced and the instant a movement is made. The time passing between these two events can be measured by letting a tuning-fork write on a revolving drum. The tuning-fork can be regulated to vibrate with great exactness, say five hundred times a second; it writes a wavy line on the drum, each undulation long enough to be divided into twenty equal parts, and thus time can be measured to the tenthousandth of a second.

The psychologist is chiefly interested in what goes on in the brain and mind. It seems that about one-half of the entire reaction-time is spent while brain changes take place, but we know very little as to these changes, or as to how the time is to be allotted among them. It is probable that in the case of the simple reaction

the movement can be initiated before the nature of the impression has been perceived. We can, however, so arrange the conditions of experiment that the observer must know what he has seen, or heard, or felt, before he makes the movement. He can, for example, be shown one of a number of colours, and not knowing beforehand which to expect, be required to lift his finger only when red is presented. By making certain analyses and subtracting the time of the simple reaction from the time in the more complex case, it is possible to determine with considerable accuracy the time it takes to perceive, that is, the time passing from the moment at which an impression has reached consciousness until the moment at which we know what it is. In my own case about sec. is needed to see a white light, sec. to see a colour or picture, sec. to see a letter, and sec. to see a word. It takes longer to see a rare word than to see a common one, or a word in a foreign language than one in our native tongue. It even takes longer to see some letters than others.

10

The time taken up in choosing a motion, the 'will-time,' can be measured as well as the time taken up in perceiving. If I do not know which of two coloured lights is to be presented, and must lift my right hand if it be red and my left hand if it be blue, I need about sec. to initiate the correct motion. I have also been able to register the sound waves made in the air by speaking, and thus have determined that in order to call up the name belonging to a printed word I need about sec., to a letter sec., to a picture sec., and to a colour sec. A letter can be seen more quickly than a word, but we are so used to reading aloud that the process has become quite automatic, and a word can be read with greater ease and in less time than a letter can be named. The same experiments made on other persons give times differing but little from my own. Mental processes, however, take place more slowly in children, in the aged, and in the uneducated.

It is possible, further, to measure the time taken up in remembering, in forming a judgment, and in the association of ideas. Though familiar with German, I need on the average sec. longer to name an object in that language that in English. I need about sec.2 to translate a word from German into English, and sec. longer to translate in the reverse direction. This shows that foreign languages take up much time even after they have been learned, and may lead us once more to weigh the gain and loss of a polyglot mental life. It takes about sec. to call to mind the country in which a well-known town is situated, or the language in which a familiar author wrote. We can think of the name of next month in half the time we need to think of the name of last month. It takes on the average sec. to add numbers consisting of one digit, and

2 In all cases the time of association only is given, the time needed to see the one word and name the other having been subtracted.