344 



NATURE 



[February 9, 1893 



bearing would falsify the direction of all that followed, 

 like a bend in a wire. 



The difificulties of dealing with detached portions of the 

 drawing, such as the eye, were easily surmounted by em- 

 ploying two of the spare letters, R and S, to indicate 

 brackets, and other spare letters to indicate points of 

 reference. The bearings included between an R and an 

 S were taken to signify directive dots, not to be inked 

 in. The points of reference indicated by other letters are 

 those to which the previous bearing leads, and from 

 which the next bearing departs. Here is the formula 

 whence the eye was drawn. It includes a very small part 

 of the profile of the brow, and the directive dots leading 

 thence to the eye. 



The letters should be read from the left to the right, 

 across the vertical lines. They are broken into groups 

 of five, merely for avoiding confusion and the con- 

 venience of after reference. 



The part of the Profile that includes U 

 &c. iiiilU jiihi &c. &c. 



Letters used as Symbols. 

 R....S = (....). Z = end. 

 U, V, T are points of reference. 



By succeeding in so severe a test case as this Greek 

 outline, it may be justly inferred that rougher designs can 

 be easily dealt witb in the same way. 



At first sight it may seem to be a silly waste of time 

 and trouble to translate a drawing into a formula, and 

 then, working backwards, to retranslate the formula into 

 a reproduction of the original drawing, but further reflec- 

 tion shows that the process may be of much practical 

 utility. Let us bear two facts in mind, the one is that a 

 very large quantity of telegraphic information is daily 

 published in the papers, anticipating the post by many 

 days or weeks. Tne other is that pictorial illustrations of 

 current events, of a rude kind, but acceptable to the reader, 

 appear from time to time in the daily papers. We may 

 be sure that the quantity of telegraphic intelligence will 

 steadily increase, and that the art of newspaper illustra- 

 tion will improve, and be more resorted to. Important 

 local events frequently occur in far-off regions, of which 

 no description can give an exact idea without the help of 

 pictorial illustration ; some catastrophe, or site of a 

 battle, or an exploration, or it may be some design or 

 even some portrait. There is therefore reason to expect a 

 demand for such drawings as these by telegraph, if their 

 expense does not render it impracticable to have them. 

 Let us then go into details of expense, on the basis of the 

 present tariff from America to this country, of one shilling 

 per word, 5 figures counting as one word, cypher letters 

 not being sent at a corresponding rate. It requires two 

 figures to perform each of the operations described above, 

 which were performed by a single letter. So a formula 

 for 5 dots would require 10 figures, which is the tele- 

 graphically equivalent of 2 words ; therefore the cost for 

 every 5 dots telegraphed from the United States would 

 be 2 shillings, or £,1 for every 100 dots or other indica- 

 tions. 



In the Greek outline there is a total of 400 indications, 

 including those for directive dots, and for points of refe- 

 rence. The transmission of these to us from the United 

 States would cost ^8. I exhibit a map of England made 

 with 248 dots, as a specimen of the amount of work in 

 plans, which could be effected at the cost of ^5. It is 

 easy to arrange counters into various patterns or 

 parts of patterns, learning thereby the real power of 



NO. I 2 15, VOL. 47] 



the process. The expense of pictorial telegraphs to 

 foreign countries would be large in itself, but not large 

 relatively to the present great expenditure by newspapers 

 on telegraphic information, so the process might be ex- 

 pected to be employed whenever it was of obvious utility. 



The risk is small of errors of importance arising from 

 mistakes in telegraphy. I inquired into the experience 

 of the Meteorological Office, whose numerous weather 

 telegrams are wholly conveyed by numerical signals. Of 

 the 20,625 figures that were telegraphed this year to the 

 office from continental stations, only 49 seem to have been 

 erroneous, that is two and a third per thousand. At this 

 rate the 800 figures needed to telegraph the Greek profile 

 would have been liable to two mistakes. A mistake in a 

 figure would have exactly the same effect on the outline 

 as a rent in the paper on which a similar outline had 

 been drawn, which had not been pasted together again 

 with perfect precision. The dislocation thereby occa- 

 sioned would never exceed the thickness of the outline. 



The command of 100 figures from o to 99, instead 

 of only 26 letters, puts 74 fresh signals at our dis- 

 posal, which would enable us to use all the 32 

 points of the compass, instead of 16, and to deal with 

 long lines and curves. I cannot enter into this now,, 

 nor into the control of the general accuracy of the picture 

 by means of the distances between the points of triangles 

 each formed by any three points of reference. Neither 

 need I speak of better forms of protractor. There is one 

 on the table by which the ghost of a compass card is 

 thrown on the drawing. It is made of a doubly refracting 

 image of Iceland spar, which throws the so-called 

 " extraordinary " image of the compass card on to the 

 ordinary image of the drawing and is easy to manipulate. 

 All that I wish now to explain is that this particular 

 application of the law of the just perceptible difference to 

 optical continuity gives us a new power that has prac- 

 tical bearings. 



Postscript. — A promising method for practical purposes 

 that I have tried, is to use "sectional" paper; that is, paper 

 ruled into very small squares, or else coarse cloth, and 

 either to make the drawing upon it, or else to lay transparent 

 sectional paper, or muslin, over the drawing. Dots are to be 

 made at distances not exceeding 3 spaces apart, along the 

 course of the outline, at those intersections of the ruled lines (or 

 threads) that best accord with the outline. Each dot in succes- 

 sion is to be considered as the central point, numbered 44 ii> 

 the following schedule, and the couplet of figures corresponding 



11 21 31 41 51 61 71 



12 22 32 42 52 62 72 



13 23 33 43 53 63 73 



14 24 34 44 54 64 74 



15 25 35 45 55 65 75 



16 26 36 46 56 66 76 



17 27 37 47 57 67 77 



to the portion of the next dot, is to be written with a firte 

 pointed pencil in the interval between the two dots. These are 

 subsequently copied, and make the formula. By employing 4 

 for zero, the signs of -I- and — are avoided ; 3, standing for — x, 

 2, for — 2 ; and i, for —3. The first figure in each couplet defines 

 its horizontal coordinate from zero ; the second figure, its ver- 

 tical one. Thus any one of 49 different points are indicated, 

 corresponding to steps from zero of o, ±1, ±2, and ±3 inter- 

 vals, in either direction, horizontal or vertical. Half-an-hour's 

 practice suffices to learn the numbers. The figures o, 8, 

 and 9 do not enter into any of the couplets in the schedule, the 

 remaining 51 couplets in the complete series of 100 (ranging 

 from 00 to 99), contain 21 cases in which o, 8, or 9 forms the 

 first figure only ; 21 cases in which one of them forms the 



