40 



NATURE 



{Nov. lo, 1887 



period ; or, in other words, as the differences of temperature be- 

 tween two short intervals that lie within the daily or yearly 

 period, minus the amount of the periodical (or normal) variation. 

 In part 4, vol. ix., of the above-mentioned Bulletin, Dr. Doering 

 has calculated the variability for Concordia (lat. 3i°25'S., long. 

 58" 4' W. ), but for three years only. The month of October has 

 the maximum value, 4° '6, and April the minimum, 2° '8. The 

 variability during spring is greatest, viz. 3° "9, and least during 

 autumn, viz. 3°'0, and the mean for the year is 3° '6, or about 

 o°'4 above that for Buenos Ayres. The hourly observations 

 published by the Meteorological Council, with the daily means 

 ready calculated, afford excellent materials for similar investiga- 

 tions. The preceding number of the Bulletin contains the 

 meteorological observations made at Cordoba during the year 

 1885. The absolute maximum shade temperature was 100° "9 in 

 December, and the minimum 14° '9 in June, giving an annual 

 range of 86° 'o. The maximum solar temperature was 147° '4, in 

 February. The mean relative humidity ranged between 817 

 per cent, in March and 61 'i per cent, in August. The rainfall 

 amounted to 24 '26 inches ; the wettest month was March, 5*96 

 inches, and the driest. May, 0*04 inch. Rain fell on 71 days, 

 and snow on one day. The times of rain at the moment of 

 observation, an element much recommended by Dr. Koppen, 

 are also quoted. 



THE WORK OF THE INTERNA TIONAL 

 CONGRESS OF GEOLOGISTS} 



II. 



M^ 



' Y only remaining subject is the representation of terranes on 

 maps by means of colours. At present no two organizations 

 and scarcely two individuals use colours in the same way ; and it 

 is probably true that every organization and individual publish- 

 ing many geologic maps has at different times employed the same 

 colour for different terranes, and different colours for the same 

 terrane. It results that the map user can gain no information 

 from the distribution of colours until he has studied the legend ; 

 before he can read a new atlas he must learn a new alphabet. 

 The advantage to be gained by substituting a universal language 

 for this confusion of tongues is manifest and great, and has 

 justified the application of much time and attention by the Con- 

 gress and its Committees. By a series of resolutions a partial 

 scheme has been selected, one colour at a time, and the com- 

 pletion of the plan has been left to the Committee on the Map of 

 Europe. That Committee has prepared a colour legend which is 

 accessible to American geologists in the volume of information 

 published by the American Committee. It is understood in a 

 general way that the Congress reserves final action, and the 

 published legend not only belongs specifically to the map of 

 Europe, but is provisional ; still, as this map, if generally ap- 

 proved, will unquestionably be declared by the Congress an 

 authoritative pattern for the guidance of map makers, the plan 

 should be freely criticized at its present stage. The selection of 

 uniform colours is a far more delicate and important matter than 

 the arrangement of taxonomic terms ; for while ill-chosen words 

 may quickly fit themselves to new uses, the adoption of an ill- 

 arranged colour scheme must entail continual loss. 



In my judgment the scheme provisionally chosen is defective 

 in several particulars, to which I shall presently call attention ; 

 but it is necessary to introduce the discussion by a statement of 

 the conditions to be satisfied by a standard colour scheme and a 

 statement of the practical means available. The following are 

 the principal conditions, arranged in an order embodying my 

 estimate of their relative importance : — 



(i) The map must be clearly and easily legible. Each colour 

 must be so distinct from each other colour that it can be identi- 

 fied, whatever its surroundings ; and all other conventions must 

 be readily discriminated. 



(2) The cartographic scheme must be adjustable to the geo- 

 logic facts ; it must not require that the facts be adjusted 

 to it. 



(3) The same scheme should serve both for general maps — as, 

 for example, those representing only systems — and for detail 

 maps, representing numerous smaller divisions. 



{4) Undue expense should be avoided. The amount and 



I Vice-Presidential Address read to Section E of the American Association 

 for the Advancement of Science, August 10, 1887, by Mr. G. K. Gilbert. 

 Continued from p. 22. 



consequent utility of colour cartography is largely limited by its 

 cost. 



(5) It should be easily fixed and retained in the mind. This 

 is best accomplished by making it orderly. 



(6) Other considerations permitting, the map should please 

 the eye. Since the arrangement of co'our areas cannot be fore- 

 told, this can only be accomplished by admitting a certain range 

 of choice. If allowed sufficient latitude in the selection of tones, 

 an expert colourist can ameliorate an offensive combination of 

 hues. 



(7) Other considerations permitting, the establishment of a 

 universal system should involve the least possible inconvenience. 

 But as the inconvenience of change is temporary, while the 

 inconvenience of a bad system is lasting, this consideration 

 should yield to every other. 



The art of mapping geologic terranes by means of colour is 

 well developed, and its methods, viewed from the geologist's 

 stand-point, admit of easy characterization. Colour may be 

 varied in two distinct ways — in hue and in tone. Hues differ in 

 quality, as yellowish-green and bluish-green. Tones differ in 

 strength, as pale green and dark green. A colour is printed 

 either solid or broken ; it is said to be broken when applied in 

 a pattern, as in lines or dots, or when it is interrupted by a 

 pattern. The difference between solid and broken colours is 

 a difference of texture. The primary discriminations in mapping 

 are through hue, tone, and texture. 



The map engraver produces texture in three ways. In the 

 first way a single impression is made with the broken colour. 

 The white of the paper, displayed where the colour is inter- 

 rupted, combines with the colour in the general effect, producing 

 a paler tone of the same hue. In the second way two impres- 

 sions are made, one with solid colour, the other with broken, 

 and the two impressions have the same hue ; they may or may 

 not differ in tone. This is monochromatic over-printing, and its 

 general effect agrees in hue with the single impression, but 

 differs in tone, being darker. In the third way two impressions 

 are made, one solid, one broken, and their colours differ in 

 hue. This is bichromatic over-printing, and its general effect 

 differs in hue as well as tone from each of the colours combined 

 in it. The first and second ways produce texture monochro- 

 matically, and do not yield a new hue ; the third way produces 

 texture bichromatically, and yields a new hue. It is practically 

 impossible to obtain a texture effect without modifying the 

 original tone. 



The natural gradation from hue to hue is absolutely con- 

 tinuous, and the number of hues is infinite ; the number of tones 

 of each hue is likewise infinite. The number of hues and tones 

 the eye can discriminate is finite, but very great ; it is stated 

 that 1000 hues have been distinguished in the solar spectrum. 

 But the number of hues and tones that can be combined in a 

 map is small. As a matter of perception, every colour is modi- 

 fied by the colours adjacent to it. The same hue affords 

 different sensations when differently surrounded, and different 

 hues may afford the same sensation. The same is true of tones ; 

 and there is a certain interdependence of hues and tones in this 

 respect. In a geologic map each colour is liable to fall into 

 various combinations, and two colours little differentiated occa- 

 sion confusion. There is therefore a somewhat narrow limit to 

 the employment of hues and tones. The matter has not been 

 fully worked out, but it is probable that twenty is as large a 

 number of hues as can safely be employed in connection with 

 tones. Texture admits of very great variation. The various 

 colour schemes submitted to the Congress and printed in the 

 report of the Bologna meeting afford, with their manifest permu- 

 tations, about 200 distinct textures, and I am satisfied from a 

 study of these and others that as many as 100 caa be chosen that 

 are not subject to confusion. It follows that a map or atlas 

 expressing few distinctions need use only hues, or only hues and 

 tones, but where numerous distinctions are to be made, recourse 

 must be had to textures. 



The printing of a large number of textures of the same hue 

 produces a greater number of tones than can be discriminated, 

 and its effect is to confiase and nullify any distinctions (within 

 the range of that hue) based purely on tone. The printing of a 

 large number of bichromatic textures causes the same result, and 

 it also produces a greater number of hues than can be dis- 

 criminated. Its effect is to confuse and nullify distinctions based 

 purely on tone, or on hue, or on tone and hue together. 



In the colour scheme prepared for the map of Europe, thirty- 

 eight distinctions are made. There are twenty-four hues, and 



