i^o 



NA TURE 



[March 31, 1910 



very small ones, and therefore showing that they 

 belong to a different category. The minus values in 

 Y are similarly treated. The process may also be 

 extended beyond the eight squares of o to loo mm. 

 in their sides, that surround the primary one. 



My experiments have been chiefly made upon the 

 "Collection of Portraits by George Dance, R.A., 

 Sketched from Life and Engraved in Imitation of the 

 Original Drawings" (Longmans and Co., 1809). They 

 were convenient to work with, being all drawn on 

 scales differing little from that of the standard. All 

 the portraits are unbearded and in exact profile, with 

 three or four exceptions. Those that are available 

 are sixty-eight in number. The name of the person 

 to whom each of the eight portraits in Fig. 5 applies 

 is written along its top, and the volume and page of 

 the two folios by Dance, from which the original was 

 traced, are given in the upper left-hand corners. 

 There are several notabilities in his collection besides 

 those in Fig. 5. Among them are Horace Walpole, 

 General Paoli, Haydn, and John Philip Kemble. An 

 exhibition of Dance's pictures was recently held in 

 London. He had a considerable reputatipn in his 

 time as a portrait painter. 



Methods have been used to aid the recollection of 

 dates and other figures. That by Gray, in his 

 " Memoria Technica," was to transform each numeral 

 into either a consonant or into a vowel or diphthong, 

 as might be the most convenient, and thereby to build 

 up words easy to pronounce and to remember. Those 

 who are familiar with such a process might apply it 

 here, and convert the four quintets of numerals into 

 four words, getting over the difficulty of employing 

 the three additional symbols as best they can. If 

 they succeed, the phrase of " four-word profiles " 

 would be literally exact. 



I do not find that a general resemblance can be 

 much increased by using one or a few more quintets 

 or words. A fifth, or even a sixth, quintet might, how- 

 ever, be usefully employed in extending the range of 

 the profile, if it contained one figure to describe the 

 chin and just below it, another to describe the brow, 

 and two figures, 00 to 99, which would perhaps suflfice 

 to give the size and general shape of the head, also 

 to define the mustachio and beard of unshaven faces. 



The next distinct stage in order of accuracy is 

 separated by a great distance from the present one. 

 It requires so large a number of dots that straight 

 or shghtly curved lines drawn through them will flow 

 smoothly when seen at the ordinary reading distance 

 from the eye. It needs as many as perhaps fifty 

 quintets to describe a profile with exactness and the 

 rest of the head with rough precision, and still more 

 to include the e5'e and ear. I have made many of 

 these, which, when reduced to the standard scale of 

 BC = 5o mm., are practically identical with the 

 originals, when viewed in a somewhat careless way 

 by a normally sighted person at a distance of 12 

 inches. A special use is made in this case of the 

 middle figure of the quintet. Thus, the numeral i 

 means that a half-unit is to be added to the first two 

 figures; 2, that it is to be added to the last two; and 

 3, that it is to be added to both of them. This power 

 of doubly minute description is often wanted in the 

 outline that joins and includes the nose-tip and the 

 two lips. Another use for the middle figure of the 

 quintet is to tell that a dotted line should be drawn 

 from the preceding point, to signify doubt of some 

 kind. A hyphen ( - ) in the middle of the quintet 

 means to begin ; an oblique line ( / ) to end ; and a 

 point ( . ) means an isolated point. But I will not 

 go further into this now ; neither will I do more than 

 hint at the way of dealing with portraits that are not 

 in exact profile, by multiplying their horizontal 

 NO. 2109, VOL. 83] 



measures into the secant of the angle through which 

 the profiles are turned away from it. 



Much more might be added on e.\tensions of. this 

 method, especially as regards its facilities and limita- 

 tions in conveying plans — ceremonial, strategic, and 

 others — for newspaper use. But its general prin- 

 ciples have been explained, and as this article is already 

 too long I will end it abruptly here. 



Francis Galton. 



TIDAL OBSERVATIONS IN THE ENGLISH 



CHANNEL AND NORTH SEA. 

 T^DR the purpose of tracing correctly the progress 

 ■*■ of the tidal wave throughout its course in the 

 English Channel and North Sea, observations of the 

 vertical movement of the tide at a distance from the 

 land, and similar observations by means of tide 

 gauges on the shore, are equally necessary. 



Information on the rise and jf^ of tide far from the 

 shore may be obtained from a ship or boat at anchor, 

 and in certain localities it may be of considerable value 

 to the navigator. 



But, if the observations are carried out with suffi- 

 cient exactitude to satisfy scientific requirements, the 

 procedure hitherto followed for that purpose necessi- 

 tates weather conditions which do not often occur, 

 and seldom last long enough for the object in view. 



There are other difficulties also to contend with, due 

 to the stream of tide running at its maximum rate at, 

 or about, the times of high and low water. The 

 stream, reversing its direction between those times, 

 causes a mark buoy, or boat, however, tautly moored, 

 to swing over a certain area during the interval ; the 

 undulating character of the surface of the ground, and 

 the action of the strong tidal stream on the lead-line, 

 thus tend to introduce elements of uncertainty which 

 increase with the depth of water. 



The practical difficulties experienced in obtaining 

 strictly accurate results by this means involve loss of 

 time disproportionate to the value of the observations, 

 and therefore the attempt has not often been made. 



Trustworthy evidence on the rise and fall of tide 

 is thus almost entirely confined to the coast-line, with 

 the exception of a few observations of tide-gauges 

 attached to the masts of wrecks on off-lying banks in 

 the North Sea. 



A large number of observations of value for the 

 reduction of soundings have, however, been obtained 

 in the North Sea during the years 1886-90 by Captain 

 T. H. Tizard, R.N., C.B., F.R.S., while command- 

 ing H.M.S. Triton in the course of the survey of the 

 shoals fronting the Norfolk coast. These, being taken 

 by the lead-line from the ship at anchor, on the as- 

 sumption that the bottom was perfectly level, can 

 scarcely be considered sufficiently trustworthy for scien- 

 tific purposes until confirmed by more precise methods 

 which were not then available. 



More rigorous observations were carried out in the 

 North Sea by the late Captain W. Hewett, R.N., com- 

 manding H.M. surveying ship Fairy in 1838 and 1840, 

 with the object of verifying the prediction of the late 

 Dr. Whewell as to the existence of an area situated 

 eastward of Orfordness and about midway between 

 the coasts of England and Holland, where the rise and 

 fall of tide was expected to vanish. 



The method employed by Captain Hewett was to 

 moor a boat head and stern as tautly as possible by 

 means of lead-lines attached to anchors laid out in the 

 direction of the tidal streams. A remarkable eleva- 

 tion in the form of a ridge on the bottom, with a depth 

 of 18^ fathoms over it, having been previously de- 

 tected, the boat was moored at slack water as nearly 

 as possible directly over the ridge. Another boat drop- 

 ping down with the tide, with lead kept just on and 



