Popular Science Monthly/Volume 63/September 1903/Palm and Sole Impressions and their Use for Purposes of Personal Identification
THE
POPULAR SCIENCE
MONTHLY
SEPTEMBER, 1903.
| PALM AND SOLE IMPRESSIONS AND THEIR USE FOR PURPOSES OF PERSONAL IDENTIFICATION. |
By Professor HARRIS HAWTHORNE WILDER,
SMITH COLLEGE.
IN a former number of this magazine (November, 1902) I gave a brief account of the epidermic ridges upon the human palmar and plantar surfaces, and emphasized their great individual difference and their applicability for use in the identification of individuals, living or dead. In the present article I shall endeavor to set forth a simple method by means of which these individual records may be formulated and classified and thus be rendered serviceable as a practical system of personal identification.
Aside from the use of photographs and the more obvious descriptive methods, which include such attributes as height, weight, color of eyes and hair, moles, birth and tattoo marks, etc., there are now in use two distinct scientific systems of identification, that of M. Alphonse Bertillon, based upon bodily measurements, and that of Mr. Francis Galton, based upon the epidermic ridges of the finger tips.
These two systems are absolutely distinct from one another, although, judging from frequent newspaper notices, they are popularly confused, with a tendency to ascribe both to Bertillon, in the same way that electrical inventions are popularly associated with the name of Edison, or theories of evolution with that of Darwin. Indeed, there seems to be a common disposition in America to ascribe the idea of the use of 'thumb-marks' to Mark Twain, who in his 'Puddenhead Wilson' has undoubtedly done much to call the public attention to the epidermic ridges of that very restricted area, although, as a consequence of the story, one continually meets with the notion that the epidermic pattern of the ball of the thumb (usually considered to be the right one alone) is individual and distinctive, while those of the remaining fingers, or the similar markings of the palm are of no importance.
It is of interest also to note that, owing to the common belief in palmistry, whereby divination is performed by means of the chance wrinkles caused by the motion of the fingers, these useless features have assumed so great importance that the far more interesting ridges appear to be usually ignored or even overlooked entirely, and as for the ridges of the sole of the foot or the balls of the toes, their very existence appears to be generally unknown.
Since there seems to be so much popular misinformation upon the subject of systems of identification, it may not be superfluous to begin the present discussion with a brief description of each of the two systems mentioned above, after which will be presented the claims of the system based upon palms and soles.
I. The Bertillon system.
The first scientific method for classifying humanity by data furnished by individual bodily peculiarities, or at least the first that became widely adopted, was that devised by M. Alphonse Bertillon, who in 1880 founded his celebrated system of identification by means of bodily measurements, 'Identification anthropométrique.' In this he applies the principles of anthropometrics, employed hitherto mainly as ethnological criteria or for use in physical culture, to the identification of individuals, using for that purpose only those measurements which depend on skeletal parts, and which are, therefore, practically unchanging after adult life is reached. The measurements selected to form the basis of his system are as follows:
| I. Measurements based upon the entire body. [Mesures relevées sur l'ensemble du corps.] |
| Standing height. [Taille—hauteur de l'homme debout.] |
| Arm reach. [Envergure des bras.] |
| Sitting height. [Buste—Hauteur de l'homme assis.] |
| II. Measurements based upon the head. [Mesures relevées sur la tête.] |
| Length of head. [Longueur da la tête.] |
| Breadth of head. [Largueur de la tête.] |
| Length of right ear. [Longueur de l'oreille droite.] |
| Breadth of right ear. [Largueur de l'orellle droite.] |
| III. Measurements based upon the extremities. [Mesures relevêes sur les membres.] |
| Length of left foot. [Longueur du pied gauche.] |
| Length of left middle finger. [Longueur du doigt médius gauche.] |
| Length of left little finger. [Longueur de l'auriculaire gauche.] |
| Length of left cubitus. [Longueur de la coudée gauche.] |
| i. e., elbow to tip of extended middle finger. |
Each of these eleven measurements is subdivided into three groups, small, medium and large [petite moyen, grand]; in accordance with definite though arbitrary limits, themselves the result of much experience ill measurements, and designed to divide any average set of measurements into three approximately equal divisions, rather than to divide equally the range of millimeters between the extremes of a given measurement. Thus, to quote an example furnished, "the numerical limits of the 'medium' head-length as used at the Prefecture of Police in Paris include an interval of but G millimeters (185-190), while those included under 'large' extend from 191 mm. to the greatest dimensions possible, an extent of more than three centimeters."[1]
Now if we were to conceive of each one of these eleven measuremenfs as varying independently of one another and as being divided into three subdivisions, the number of possible subdivisions under which individual anthropometric records could be filed would reach the large number of 3 to the 11th power, or 177,147, but in practical use M. Bertillon employs for purposes of identification only a few of these measurements, which he gives in the work just quoted, together with an hypothetical application, as follows:
He supposes the case of 90,000 sets of measurements, a number approximately corresponding to that of the adult male prisoners recorded in the Paris prisons up to 1893. Of these the first classification is made by means of the length of head, and as the subdivisions, small, medium and large, are fixed with reference to equality of division, approximately 30,000 of these records will be placed in each.
Each of these subdivisions is now divided again into three parts, in accordance with the breadth of head, a division which leaves approximately 10,000 in each of the nine compartments, i. e., 10,000 individuals whose head length and head breadth fall into the same categories. The third division, which reduces the number of records in each of 27 compartments to about 3,300, is based upon the length of the left middle finger, and the fourth, resulting in 1,100 in each of the 81 compartments, is based upon the length of the left foot. The length of the cubitus then follows, which increases the number of compartments to and the number of individual cases in a compartment to less than 400. By the addition of the standing height, the number of compartments is increased to and that of the cases in each compartment to approximately 130, and these numbers become respectively and 42 by the use of measurements taken from the left little finger. These are finally reduced to small sets of a dozen records each by such criteria as the color of the eye and the length of the right ear, after which this small number may be carefully compared for individual measurements.
With some modifications the above system is in official use in most of the civilized countries of the world, including England, Russia, Belgium, Switzerland, the United States and the majority of the South American republics, but in at least some of these cases the governmental acceptance of the system does not mean an extensive practical use, a condition of affairs especially true in the United States, where the governmental acceptance means merely an official sanction and where each state, or even each municipality, may employ its own methods of recording and registering criminals, quite independently of its neighbors. In this country the main reliance is placed upon photographs and descriptions, sent to various police headquarters in the form of little handbills, and although one sees occasionally among the descriptive part of these a set of Bertillon measurements (without further designations), it is very doubtful if in cities of moderate size the police authorities have any definite idea of their significance, or possess the necessary instruments for obtaining these measurements and thus verifying the data furnished. In England the Bertillon system is extensively employed, although in a somewhat modified form, to which is appended, at present as a supplementary system, that of Galton, to be described below. On October 21, 1893, a departmental committee was appointed by the home secretary, the Hon. H. H. Asquith, to inquire into the various methods of identification of criminals, and an official report was presented by them on February 12, 1894, and published as a Bluebook (C. 7263). The recommendations embodied in this report, and adopted in full by the English government, were as follows (paraphrased):
| I.To photograph criminals as at present, the photographs to consist of both a front and profile view, taken on separate negatives, and not by means of a mirror, as heretofore. |
| II. To employ the first five of the Bertillon measurements, as follows, expressed in millimeters: |
| 1. Length of head. |
| 2. Breadth of head. |
| 3. Length of left middle finger. |
| 4. Length of left forearm. |
| 5. Length of left foot. |
| III. To take the finger prints by Mr. Galton's method. |
| IV. To add a brief description including the height in feet and inches, color of hair, eyes and complexion and distinctive marks, the latter in a fixed order, beginning with the head, then the hands and arms, then the body, and lastly the legs and feet. |
With regard to that portion of the recommendation which concerns the Bertillon system the committee gave its unqualified approval to the use of the first five categories as given above, but felt that the further subdivisions (height, length of little finger and color of eyes) were rather unsatisfactory. As stated in the report:
For the 'primary classification' that based on the first five Bertillon measurements, a complete outfit, such as would be necessary at an important registration station, would consist of 343 drawers, corresponding to the 5th power of 3, the number of possibilities involved.
The above quotation from the English report is given in full mainly for the purpose of showing the great disadvantage to the entire system, which results from racial differences in bodily proportions, a fact which will necessitate either one of two alternatives, both bad; that of using special figures for each country or of having very unequal subdivisions in certain cases. This is a decided barrier to the internationalizing of the system and must necessarily be reckoned as a serious defect.
Without meaning to seem ungracious to a system the advantage of which over all previous methods has been universally recognized, and one the scientific principles of which reflect so much credit upon the deviser, it may be well, in closing this brief account, to enumerate the defects of the Bertillon system, some of which are, indeed, incident to any system which human ingenuity can devise, and the most of which have been foreseen, acknowledged and corrected so far as possible by M. Bertillon himself.
1. The limitation of the system to the period of adult life.
2. The necessary disparities between the same measurements taken at different times by different mensurators, or indeed by the same one (percentage of error).
3. For the purpose of an equal classification, the necessity of assigning independent limits to the records kept by each nation.
4. The greater amount of time consumed in making a set of measurements and in recording and classifying the same, as compared with the printing and reading required by either the Galton system or by the one advocated in the present paper. A careful test of this has not yet been made, but when we consider the number of single acts involved in the making of the records required by each system, the conclusion is obvious. In identifying an individual by means of a previous record, the Bertillon system demands a complete remeasurement, while by the palm and sole system a mere glance at a single palm is often sufficient to establish the identity or the reverse. Ordinarily the difference of time may not be great, but in the stress of modern competition a slight disadvantage in this particular may be regarded as a relative defect.
5. A more serious defect, which is also brought out by comparison, is that the certainty of a Bertillon determination is not absolute, while that of a system which involves either the finger tips or any other considerable portion of the epidermic ridges of hand or foot is beyond question. This has been thoroughly proved statistically by Galton and morphologically and embryologically by a series of recent investigations in my laboratory.[2] The proof afforded by the study of duplicate or 'identical' twins, where the resemblance, though greater than it can be in any other two persons, is still not absolute, affords farther evidence of the same.[3] Galton says that a proved identity of finger prints "far transcends in trustworthiness any other evidence from any number of ordinary anthropometric data. By itself it is amply sufficient to convict. Bertillonage (i. e., the system of Bertillon) can rarely supply more than grounds for very strong suspicion; the method of finger prints affords certainty.[4]Although in the original system devised by him Bertillon confined his attention mainly to anthropometric measurements and rejected all use of epidermic marking of hand or foot as impracticable[5] in his capacity as chief of the Bureau of Identification and with the evident desire of bringing his work to the highest degree of efficiency, he has recently adopted a part of Galton's system, and places the impressions of certain of the finger-tips upon his identification-cards.
Unfortunately, I can not ascertain the exact date at which this adoption of Galton's system was made, but upon a fac-simile card, shown in a recent popular article on Bertillon's system (Leslie's Weekly, April 16, 1903), which is dated August, 1901, spaces appear below the words 'Pouce, Index, Medius, Annulaire' and are plainly intended for the reception of the corresponding finger-prints. Within a few weeks of the present writing there have appeared in various newspapers (e. g., Boston Herald) accounts of the employment in the State Prison at Auburn, K. Y., of imprints both of the fingers and of the entire palm, but I am unable to ascertain anything definite concerning the manner in which these prints are to be used. Mr. John ]Sr. Ross, the chief of the Bertillon department of the above-named prison, has kindly given me what information he can concerning the matter, but writes that it is 'an entirely new departure' and that 'directions as to its application have not yet been received by the Bertillon operators of the different penal institutions' (June 29, 1903). I am thus unable to say whether M. Bertillon has in mind the incorporation of any part of my system with those of himself and Mr. Galton, but I have furnished him with reprints of my two previous papers on the subject and have sent him also numerous manuscript notes and sample prints, which together present the essential points of the system as given in this paper.[6]
Should this system of mine be found of value and permanently incorporated with the others, the 'Bertillon' system known in actual practice will be, like most other inventions of real value, a composite resulting from the independent investigations of several individuals working from different standpoints, and should be carefully distinguished from the real Bertillon system as described by him in his published work, and outlined above.
II. The Galton system.
In the popular mind, as attested by numerous works of fiction and by newspaper articles, the main use of the patterns of the finger tips is to aid a detective in identifying a criminal by means of the marks which his fingers have left upon the objects which he had handled; but, as a matter of fact, although such a proceeding is certainly possible, Galton seems never to have suggested such sensational aid to detective work. Both the Bertillon and the Galton systems are rather methods of describing and registering a man, whether a criminal or not, by certain physical peculiarities and in such a way that he or his body may be identified at any future time; and both involve two procedures, (1) that of taking certain individual records, and (2) that of classifying and arranging them so that they may be easily found when occasion requires. The Galton system is based upon imprints of the epidermic patterns found upon the balls of the thumbs and fingers, and Mr. Galton, although by no means the first to employ such means for the identification of individuals, is the first to attempt a careful and scientific system by which these data may be described, registered and classified.
The use of such prints has been sporadically employed in both ancient and modern times, and seems to have long been in use among the Chinese, but data concerning the official workings of this vast and ancient empire are difficult of access to Europeans, and it is likely, as in so many other claims, that the facts when found will be disappointing when compared with the reports concerning them. Galton himself was first led to the study of finger prints by his friend Sir William Herschel[7] who, when 'Collector' or chief administrator of the Hooghly district in Bengal, added to the signatures of the natives upon all official documents the imprint of the index and middle fingers of their right hands, taken by means of the ink employed for his office stamp.
Galton, indeed, says of his friend that 'if the use of finger prints ever becomes of genuine importance, Sir William Herschel must be regarded as the first who devised a feasible method for regular use and afterwards officially adopted it,' but it must also be remembered by the one who writes the final history of this system that it was Galton who devised the method by which such prints could be described and classified, and thus become of practical value.
The Galton system of personal identification by means of finger prints rests upon two necessary principles, both of which have been established by him beyond refutation:—
I. The absolutely individual character of the markings.
II. Their permanence throughout life.
The records employed are the printed impressions of the ten digits placed in a definite order upon a card, and the separate cards are placed on file by means of a classification wholly dependent upon the individual patterns.
Of those latter there are three types, the arch, the loop and the whorl, designated in descriptive formulæ by their initial letters. A, L and W (hence the name of 'the alw system' by which Galton has designated it). Of these the loop, which may turn to either the radial or the ulnar side of the hand, is for some purposes farther subdivided into radial and ulnar loops designated respectively as R and U. The patterns are definite in their nature; transitions between them are of rare occurrence and these are nearly always referable to one type or the other. This system of 'ALW,' with the occasional subdivisions of the R and U, forms what Galton designates the 'Primary Classification,' and the finger tip records of any number of individuals are arranged in accordance with a preconceived order. Although he has made several experiments in this, in his final method (1895) the sets are first arranged in four divisions (ARUW) in accordance with the type of pattern found upon the right index finger. This is followed by (small) letters designating the patterns upon the middle and ring fingers of the same hand, using l instead of r and u for all looped patterns. This will be seen to subdivide each of the first four divisions into 9 or will divide an entire set into 36, as follows:
| Aaa | Raa | Uaa | Waa | |
| Aal | Ral | Ual | Wal | |
| Aaw | Raw | Uaw | Waw | |
| Ala | Rla | Ula | Wla | |
| All | Rll | Ull | Wll | |
| Alw | Rlw | Ulw | Wlw | |
| Awa | Rwa | Uwa | Wwa | |
| Awl | Rwl | Uwl | Wwl | |
| Aww | Rww | Uww | Www |
If, now, the prints in each of these 36 divisions be farther subdivided in accordance with the same three fingers of the left hand, each subdivision consisting likewise of 36 compartments, the entire collection will be divided into 362 or 1,296. The thumb and little finger of the right hand which show 9 possible combinations, will subdivide each of the 1,296 compartments and, in like manner, the patterns of the thumb and little finger of the left hand will give another subdivision of 9, so that the number of possible compartments or subdivisions into which a set of prints may be arranged by means of this primary classification alone is 1,296 × 9 × 9 or 104,976; and since the various combinations with a few exceptions occur with about the same frequency, the number of separate prints in a collection of five hundred thousand which it would be necessary to look over carefully in comparing with a certain definite case would average about five.
The data necessary for this latter comparison are abundantly furnished by the details of the individual ridges, termed by Galton the 'minutiæ' and the farther description and subdivision of the records by means of these he terms the 'Secondary classification.' These and other useful details are appended to the formulæ given above by means of 'descriptive suffixes' arbitrarily selected and described in a table, a copy of which must, of course, be always at hand, at least until it be thoroughly committed to memory by the clerk in charge of the records. The nature of these suffixes and of the details which they describe may be learned from the following examples, taken at random from Galton's table:
| g. | The core to the whorl is very large. |
| o. | The core of the whorl is a detached ring. |
| x. | Interpretation questionable; the pattern is peculiar. |
| † | Scar left by a cut. |
The implied suggestion of 'x' brings up a question which probably occurs to the reader at about this time, namely, whether a pattern is ever of a mixed type, or half way between two, thus giving a chance for a difference of interpretation and a consequent embarassment in finding the case from the formula. This certainly occurs occasionally, but Galton has well disposed of the difficulty by comparing it to the doubt experienced when consulting a city directory for a Scotch name beginning with 'Mac,' variously written, either in full or as Mc or M', and classed differently by various lexicographers. In both cases the investigator, failing to find what he desires in one place, looks in another, and neither here nor there is the difficulty a serious one.
Each finger tip record is placed on a card measuring 12 x 5 inches, and contains, when complete, rolled impressions of the ten digits, a set of 'dab' impressions of the four fingers of each hand (as duplicates for comparison) and, at the right hand upper corner, the formula. In Scotland Yard a folded paper is used instead of a card, and the arrangement of the prints differs somewhat from the above.
Concerning the practical adoption of the Galton system at present, it is hard to get details, but the recommendation of the English committee in 1894 has been referred to, and it seems that since that time the method has come into quite general use in England. It is of course impossible to devise a method which in every detail will be perfect from the start, and as Mr. Galton is continually at work upon his system, the improvements suggested both by him and by those practically engaged in the work can not fail to modify details until it is brought to the highest degree of efficiency.
III. The palm and sole system.
The method which it is the purpose of this paper to advocate and briefly explain is closely allied to that of Mr. Galton and is, in fact, an extension of his system to the palmar and plantar surfaces, which are covered with the same sort of ridges as are the finger tips and in which the variation is greater and the details larger and more obvious. A moment's inspection of a human hand and foot will show that the entire ventral surface of each, including that of the digits, is covered by a peculiar sort of skin, very different from that found elsewhere, and that along the sides of the palm and the digits, and just above the sole upon the foot, there are definite lines of separation between them and the normal skin, which in the hand corresponds in general position to the seam in a glove which unites the upper and under surfaces. This palmar and plantar skin differs from that of the rest of the body in many ways. It is absolutely hairless and at no time during embryonic life shows indication of either hairs or hair follicles. It consequently has no power of forming goose-flesh when chilled, although the back of the hand and the surface of the forearm, in the immediate vicinity of the palm, are favorite places for the display of this phenomenon. It is also very slightly, or not at all, pigmented as is readily seen by inspection of the palms and soles of a negro, and consequently does not tan or freckle, a distinction often made very obvious by a comparison of the back of the hand with the palm. The most obvious character, however, and the one which directly concerns both Galton's system and the one advocated here, is that of the small but distinct epidermic ridges, which cover the surfaces in question. These may be said to run in a general way parallel to one another and diagonally across the palm or sole, although in certain regions their direction is altered and at more or less definite places they form curiously disposed patterns, usually in the form of loops or spirals. With a moderate lens these ridges give the skin an appearance much like that of corduroy and there may be seen running along the middle of each ridge a row of minute indentations or pores, at about equal distance from one another, the orifices of the perspiratory glands. Running over and across these ridges in directions which bear no relation to them are the wrinkles or rugæ, more abundant in the hand than in the foot, and caused by the various motions of the digits and of the other movable parts of the member. Those seem at first especially obvious and interfere more or less with the study of the ridges, but a little practice will enable one to ignore them altogether. In printed impressions, which are used for purpose of study far more than are the actual surfaces, most of these are pressed out of existence while the remainder appear merely as narrow white streaks which do not affect the investigation (see Figs. 1 and 2).
These ridges and their peculiar disposal are an inheritance to us from our arboreal ancestors, and appear to be formed in the oldest primates by the coalescence of single units which arrange themselves in rows.[8] Whether or not this phylogenetic or racial stage is now passed through in each human embryo in accordance with the law of biogenesis has not as yet been shown, but it is certain that the ridges are seen fully formed and in their adult condition in a four-months' embryo, and that no change can afterwards take place in any detail.
As these surfaces are thus individually variant and as their condition is absolutely permanent throughout life, they offer the best possible criteria for a system of individual records, especially since they may be so easily recorded by means of printed impressions. All these points have been shown in a practical way by Mr. Galton, who has taken as the basis of his system the markings that cover the balls of the fingers, his 'finger-tips.' The present paper considers the remainder of the ridged surfaces and is thus seen to be an extension of the Galtonian system to a new territory. Whether ultimately the universal personal records, which will surely become a necessity in the near future, will be based upon a part or the whole of these surfaces is of no real moment and it is with the idea of being of genuine assistance to Mr. Galton and without any attempt at rivalry that I offer in the following pages a method of recording identity by means of palms and soles.
M. Bertillon has said that there are not lacking individually variant parts of the body capable of use for purposes of identification, but that what is needed is some system of recording and classifying these differences, so that an individual case can be easily found. The system proposed here will, I think, fulfil this demand, and it will be seen that each human being is as well marked and labeled as though he were tattooed with an individual name and number, the interpretation and manner of cataloguing these devices being the only part not furnished by nature.
The method of printing a palm or sole is a very simple one, and although there are many little details which will occur to one who does much of this work, the essentials are the same in all cases. The outfit for printing consists of a tube of mimeograph ink, a rubber roller such as is used in amateur photography, and unruled paper of the required size. An inking surface is prepared by pinning a sheet of paper to a board or table placing upon it a little of the ink, and then rolling it down with the roller until the paper is coated with a uniform thin layer of ink. The best results are obtained when this is thin enough to appear of a dull green color rather than black, and the usual difficulty lies rather in using too much than too little ink. The hand or foot to be printed is then laid upon the inked surface, pressed a little, especially at the places which are naturally raised above the paper, and then removed and laid in the same way upon a clean sheet of paper, pressing the parts as in the first instance. Care must be taken not to slip the hand or foot sideways at any time, as this would blur the lines; a similar condition may be caused by too great pressure, and thus when the feet are taken the subject should be seated, allowing the foot to be manipulated by a second operator. The inking surface should be freshly rolled before each new impression, and when a number are taken at one time, a little ink must be occasionally added and rolled down. If mimeograph ink is not available, ordinary printer's ink will do almost as well, and both sorts may be readily removed from the skin by the use of a little turpentine or benzine, or even by soap and warm water.
After a collection of imprints has been made, the next procedure is the interpretation, that is, the tracing out of certain definite lines which mark the course of the ridges and define the patterns. As the palm presents simpler conditions than does the sole and is much the best for purposes of instruction, we will begin with a good average print like that given in Fig. 1, using a sharp-pointed pencil, and, when necessary, a reading glass of low power. At the base of each of the four fingers there will be seen a triangular area composed of transverse ridges, so intruded into the palm that it parts for some little distance the ridges which belong more definitely to the palm itself. These are the four digital areas, and at their apices are found points from which the ridges radiate in three directions, two bounding digital areas and one traversing more or less of the palm. These four points or triradii (equivalent to Galton's 'deltas') are the starting points of the system and may be termed the four digital triradii, numbered 1-4, beginning at the inner or thumb side. The lines bounding the digital areas are the eight digital lines, numbered from 1-8, and the four other lines which proceed from the triradii and cross the palm are the four main lines. These latter, designated by the letters A-D, are of primary importance and furnish by their course the first or primary formulæ by which the palms are classified. These lines are established by following the direction of the ridges to whatever point they may lead, and are best traced along a certain definite ridge, although, in places where a ridge that is being followed breaks or forks, the line should be continued by means of an adjacent ridge, or by taking the general direction indicated by several ridges. When the four main lines are traced, search should be made for a fifth triradius, the carpal, usually occurring in or near the middle of the palm just above the wrist, and its lines should be followed in the same way as in the other cases. The two short lines running downward to the wrist are the carpal lines and define the carpal area, and the longer one which curves about the base of the thumb is the thenar. A carpal triradius is not always present, but in some cases its place is taken by what may be termed a 'parting,' or a place where the ridges which run from the palm to the wrist divide
into two groups, one half of which diverge toward the inner, while the other half diverge toward the outer side (see Fig. 4, a and b). When the interpretation is complete the palm impressions will resemble Fig. 2 (compare with Fig. 1).
If, now, similar impressions are made of a half dozen palms, a great individual difference in the course of the main lines will be at once apparent. They will curve in different directions, sustain various relations with regard to one another and terminate at different points along the margin of the palm. Although occasionally two palms will show the same general course of the main lines, there are, on the other hand, a large number of distinct eases, and this system may well serve for a primary classification, or one which will divide individual records into a large number of sets, as described above in the case of the other two systems. In the former article in this magazine, I suggested this by applying names to the various areas marked off by the lines of interpretation, and proposed a set of descriptive formulæ based upon these areas to be used in designating the course of the lines. I now wish to substitute for this a numerical system, the presentation of which in
the form of a key explains itself (Fig. 3). In this, each triradius and intermediate area is furnished with a number, the latter being designated by the odd, and the former by the even, numbers, and the course of a given main line may he simply and accurately described by giving the number corresponding to the point at which it terminates. To the extensive outer border where the use of a single number would be often indefinite, three numbers are assigned, 3, 4 and 5, although where complete accuracy is not needed the symbol (for open) may be used, signifying merely that a given line passes out at some point along the free margin between the outer carpal line and the outer digital one of the little finger. Of these three numbers, 4 is used to designate a pattern only, which occasionally occurs in this region, and which, when present, defines the territory below it as 3 and that above it as 5. A line entering it and returning along its lower side would be designated as 4/3 (Fig. 4, d), one emerging above as 4/5 and one that becomes involved in the pattern and does not emerge is simply 4.
Fig. 3. Diagram of a Left Palm, showing the designations to be used in making the descriptive formulæ. Compare with Fig. 7. When a pattern is not present in this region, 3 may signify approximately the lower third of the entire outer margin, and 5 the remainder.
In the majority of cases a main line will terminate in one of three ways, it will either (1) open freely along the outer margin; (2) cut through an interspace between fingers or (3) it may fuse with another main line, forming a loop. In this latter case each of the two lines involved may be considered as terminating in the triradius of origin of the other and be described by the corresponding number. Aside from these three possibilities, there is an occasional fourth one produced by the presence of what may be designated 'lower triradii.' So far as has actually been observed (in above 600 hands) these may occur in but two places, as designated in the diagram (Fig. 3) by inverted 
s in the 2d and 4th of the digital interspaces (counting that between the thumb and the first finger as the 1st), but there is morphological ground for expecting one also in a corresponding position in the 3d interspace. When a main line terminates in one of these triradii it should be designated by a , with an exponent signifying the space to which the belongs; thus 7 or 11 as the case may be. In addition to these, pattern triradii may occur in connection with the thenar and hypothenar patterns, and of these the second, 11, comes into occasional relation with line A.
Having now a method by which the course of the four main lines may be designated by means of a sequence of four figures, let us illustrate this by a few cases taken at random, and represented by the tracings given in Fig. 4.
For these the main line formulæ will be as follows:
(a)5.7.9.11.
(b)5.5.9. 9.
(c)2.7.8. 11.
(d)436.9.10.
In these the third line of (c) is designated by 8, the number for its point of origin, since it exhibits a course not unusual for it, but never found in the other lines, that of running down into a loop and not emerging, so that it can not be said to have a point of termination. In order to obtain formula? enough to work with, we may add to the above that of the example (Figs. 1 and 2) 2.5.7.9; also those formulæ designating the four palms figured in Fig. 6 of the previous article, viz: (a) 5.5.5.7; (b) 3.5.6.8; (c)45.6.7.10; (d) 5.8.10.11.
Fig. 4. Tracings of Four Left Palms, showing various line formulæ. (a) 11.9. 7.5 [Collection No. 109]; (b) 9.9.5.5 [Collection No. 323]; (c) 11.8.7.2 [Collection No. 30]; (d) 10.9.6.43 [Collection No. 32]. In (a) there is a parting in place of the carpal trirtidius; in (b) there is a well-developed thenar pattern; in (c) line C is very short and runs into a loop where it ends abruptly; in (d) the third lower triradius is present, assisting in the formation of a pattern.
To arrange these or any number of formulæ in definite order it will be necessary only to make the first subdivision in accordance with the first designation (i. e., that of line A) and so on with each designation in succession, employing the usual numerical sequence. A fraction may be marked by its numerator alone, since the denominator is nothing more than an added specification or descriptive mark, and the sign L may take precedence of all, ranking before the figure 1. By this means the nine formulæ referred to above would be arranged as follows:
| 2.5.7.9. | 45.6.7.10. | 5.5.99. |
| 2.7.8.9. | 43.6.9.10. | 5.7. 9.11. |
| 3.5.6.8. | 5.5.5.7. | 5.8.10.11. |
Although, as given in all of the above illustrations, the natural order of sequence in the four designations of each formula is from the line A to line D, it happens that of the four, A is the most uncertain in its interpretation, and is the only one concerning the designation of which differences of opinion would be likely to occur. It is unfortunate to begin with this especial one and thus be liable to be put on the wrong track at the outset, and it appears to be better to reverse each formula, beginning with line D. and ending with A. This would lead to a rearrangement, not only of each formula which would be simply reversed, but also of the order of sequence of the whole, so that the nine formula used above would become rearranged as follows:
| 7.5.5.5. | 9.8.7.2. | 10. 9.6. |
| 8.6.5.3. | 9.9.5.5. | 11. 9.7.5 |
| 9.7.5.2. | 10.7.6.45. | 11.10.8.5 |
in which a given formula would be as easily found as in the other order, with the advantage of having the uncertain designation in the fourth place instead of in the first.
Of course it can not be known, at least as yet, what is the total number of main line formulæ which occur in the human hand, but in this regard the following table will be of interest, which gives in the regular order of sequence the formulæ of the hands of 100 American female college students, 100 rights and 100 lefts, and shows the number of times each formula occurs in each hand.
Table I.
By this it will be seen that the total number of formulæ represented in the 200 hands is 61, that there are 48 separate formulæ in the left hands and but 38 in the right. Of the 48 left hand formulæ 23 do not occur in the right, and of the 38 right hand formulæ 13 are not found in the left, 25 being common to both. The commonest formula for the left hand is 8.6.5.3., which occurs in 9 per cent, of the cases, while for the right hands the commonest formula is 11.9.7.5., occurring in 22 per cent, of the cases. The records from so small a number of cases can not be conclusive, but it would seem both by the number of formulæ represented and by the smaller percentage of occurrence of the commonest formula that the left hand is much more variant.
From this table there may be also calculated the amount of latitude allowed in the positions assumed by each line and the percentage of occurrence of each terminus of each, the results of which are given for convenience in a separate table, as follows:
Table II.
It will be seen by this that the commonest position for line D is 11 (between the index and middle fingers), a position which occurs in 75 out of 200. Similarly the commonest position for line C is 7; for line B, 5; and for line A also 5; but, curiously enough, the commonest actual formula is not 11.7.5.5., the combination of these. This table also shows that line D may oscillate in its position from the interspace between the ring and little fingers to that between the middle finger and the index; that line C swings between an open position and the same limit as that of line D, and so on.
In the employment of these line formulæ as a primary classification it seems advisable for several reasons to employ that of the left hand first, which will be seen to divide a series into between 40 and 50 compartments, and since the right hands appear to vary independently or nearly so, the addition of the line formulæ of those would subdivide each of the 40 or 50 into about the same number of lesser compartments, or in all approximately 452 or 2,025.
That is, if the hand-prints of a city of 100,000 inhabitants were arranged in accordance with the line classification alone, there would be needed over two thousand compartments, with, theoretically, about 50 in a compartment. As practically worked out, the distribution of these prints in compartments would be an unequal one, and while certain of the compartments would have but a single representative, or be even empty, some of the others would have a much larger number than fifty, possibly even several hundred.
But the line formulæ are but a primary classification and do not by any means exhaust the resources of these very varying parts. As a secondary classification, for the farther subdivision of the palms, to be appended to the first and employed whenever expedient, the use of a pattern formula may be suggested. This is based upon the irregular occurrence of patterns such as occur normally in the simian hand and appear sporadically upon various localities of the human palm.
If we count as a pattern each place where there is a definite loop, whorl or noticeable disturbance of the usual even course of the ridges, although by doing so we violate certain of the underlying morphological principles as shown by the comparison with other mammals, we may expect to find any or all of five patterns, located as described in the previous article and corresponding roughly to the divisions into which the palm is divided by the lines of interpretation, viz., one thenar, one hypothenar and three interdigitals (palmar of the previous paper). Of these the thenar, of rare occurrence among people of the white race, is morphologically composed of the vestiges of two, a genuine thenar and the interdigital which corresponds to the interval between the thumb and index; the three interdigitals lie below the three intervals between the fingers from the index to the minimus; and the hypothenar appears upon the large eminence of the same name which forms the outer boundary of the palm.
A pattern formula will thus need five places, one for each of the five patterns in a prearranged order, and, again adopting the principle of placing in the first position a very obvious one, about which there is no doubt, the order suggested may be, hypothenar, thenar, first, second and third interdigital. When any one of these is present, it may be designated by an abbreviation, such as a capital H for hypothenar, Th (or better, the Greek θ) for the thenar, and 1, 3 and 3 respectively for the others. When absent, may fill the position, and when there is merely a rudiment, a small r may be added to the as an exponent. Thus a few representative pattern formula" would be the following, taken from actual cases:
| H.0.1.2.3. | 0.0.0.0.3. |
| H.0.0.2 3. | 0.0.1.2.3. |
| H.0.0.0.3. | 0.0.0.2.3. |
| H.0.0.0 0. | 0.0.0.2.0. |
| 0.0.0.2.0. | 0.0.0.0.3. |
| 0.0.0.2.0. |
The order of arrangement for each position would be naturally to consider the signs of the patterns as precedent to zero, an arrangement which it will be noticed, has been followed in the above list. A rudiment, like Galton's descriptive suffixes, is disregarded in the arrangement, and counts like other zeroes.
Fig. 5. Print of a Right Sole [Collection No. 112] showing a High Degree of Complexity, and presenting difficulties in formulation. There are two lower triradii, probably the first and second, in which D and B [IV and II] terminate respectively: the triradius of line C[=III] is beyond the limit of the print and its location is in part conjectural; the second and fourth digital lines curve downward across the palm.
Morphologically the same pattern, i. e., one in a given position, may differ considerably in regard to its mode of formation, the presence or absence of 'pattern triradii' or those concerned in its structure, the shape of the pattern itself, its degree of completeness and so on, and all these attributes may be easily added by means of a series of easily devised descriptive signs, like the 'descriptive suffixes' of Galton, and used as exponents, having, like the exponent r, no influence upon the arrangement of the formulæ, but simply completing the description.
Should a third classification be necessary, the presence, position or absence of the carpal triradius would furnish one, and for a fourth subdivision a counting of certain of the ridges, like those between the digital triradii as Galton does in his finger-tip system, might be suggested. In the last instance, where the decision of a definite case rests upon the identity of a certain individual, actual prints taken from his palms should be compared with the set in the collection which most closely corresponds in the formulæ to them, and the decision should be reached by the study of the minutæ, although there could hardly he a mathematical chance of a comparison going as far as that unless it was a case of actual identity.[9]
Fig. 6. Tracings of Three Left Soles, showing various relations of digital and main lines, (a) Lines A, B and C are open; line C is recurved and opens at 9; the third digital line is recurved around a pattern: the first and second digital areas are confluent; there is a single lower triradius [Collection No. 156]. (b) Lines A and B are open; line C is recurved; line D terminates in a lower triradius (the third?); the digital lines are all normal and the digital triradii are distinct; there are two lower triradii [Collection No. 86]. (c) Lines A, C and D are open; line B is recurved around a pattern and fuses with one of its own digital lines; the digital lines are in general much modified and the digital areas are all confluent; there is but one lower triradius, the first [Collection No. 59].
The previous pages have treated of the hands alone as though they were the only possibility of the system, but the soles of the feet exhibit fully as great a diversity in the course of their ridges, and their use in addition to that of the hands would furnish so complete a means of subdividing a collection of prints that the secondary classification, i. e., that which concerns the patterns and the carpal area, would hardly be necessary except to complete a description. The feet as well as the hands possess four main lines originating from the digital triradii, and as their formulæ are nearly as variant as are the others, the subdivisions rendered possible by the use of all four members are well-nigh infinite in number (see Figs. 5 and 6).
Thus if, as shown above, a set of formulæ would be divided into upwards of 50 divisions by using the left hand alone, and if each of these would be farther subdivided by adding the formulæ of the right hand, producing 2,500 divisions in all, the addition of the left foot to these might increase the number to 2,500 X 50, or 125,000, and these would become 7,250,000 by the use of the right foot, or enough to characterize every citizen in a large state or small country, employing merely the primary classification. It must be remembered, however, that these are theoretical figures and that the actual combinations of lines may not be as great, nor would the various kinds be as regularly distributed; yet enough has been shown to prove that the number of separate actual combinations of the line formulæ alone, if both the hands and feet are employed, would be very great.
In a sole print the characteristic features are mainly distributed along the ball of the foot, anterior to the hollow of the arch, and while in a general way they are similar to those of the hand, there are also numerous important differences, some of which will be seen in Fig. 5, a print which represents a more complex condition than is usually seen, and in the tracings given in Fig. 6. The four main lines of the sole, although arising from digital triradii, usually curve towards the inner instead of the outer side, and when open, are apt to converge at the inner margin almost or quite to the point of fusion. There is also almost always upon the thenar region or ball of the great toe a conspicuous pattern, which may be termed the hallucal pattern. This possesses one or two, and possibly three triradii, of which the upper one is the proper digital triradius of the great toe, usually unrepresented in the hand; while there is often a second one upon the extreme inner margin, sometimes shown only by rolling the foot a little during printing. The hallucal pattern shows much variation and is easily divisible into a series of types, which well serve the purpose of a secondary classification. Lower triradii are of far more frequent occurrence than in the hand, and are often located so near one another through the convergence of the interdigital areas that it is difficult or impossible to attribute them to any one of them.
Probably the greatest barrier to the formulation of sole conditions in the same way as in the case of the palms lies in the position of the digital triradii, which are apt to be situated in the hollow beneath the toes and thus beyond the margin of a print—a condition especially apt to occur in the case of the third one, i. e., that at the base of the fourth toe; again, the relationships of the triradii are often complicated by the fusion of two or more digital areas with one another and the consequent displacement of the digital lines, which may simply pass one another upon the digital areas or curve downwards over the ball of the foot as is the case with the second and fourth digital lines in the print given here (Fig. 5).
These difficulties, especially that of the extra-limital position of the digital triradii, certainly prove a barrier towards the application of the same system as in the palm; yet, with some adjustment to conditions, this would still seem to be feasible. A diagram assigning numerical designations to the 
Fig. 7. Diagram of a Left Sole, showing the designations to be used in making the descriptive formulæ. Compare with Fig. 3. different terminal regions is given here (Fig. 7) which may be compared with the similar one referring to the palm (Fig. 3).
A more detailed account of the sole formulation is not within the space limit of the present article, but enough has been given to suggest how this may be accomplished. I would especially emphasize the practicability of the use of the hallucal pattern, perhaps even in a primary classification, which recommends itself by its large size, its conspicuous character, and its ready divisibility into definite types.
Lastly, there remains only a short discussion of the means of recording and filing away prints, the amount of space they would occupy and their consequent feasibility as a means of recording all citizens, as advocated in the previous article. The prints themselves should he taken upon smooth but not glazed unruled paper of a suitable size, one of of about 35 X 21.5 cm. (14 x 81⁄2 inches) for the hands, and one of 35 x 28 cm. (14 x 11 inches) for the feet. Upon each of these the respective prints should be arranged in the natural order, that is, the left upon the left side, etc. There is enough blank paper left in such a piece to record any details necessary about the person; the name, date of birth, nationality, and even the Bertillon measurements.
As in the Galton system the formulæ may be written at one end, most conveniently the left, so that an especial case could be found by turning over a pile of papers as in selecting a page in a book; or else each set of prints may be simply numbered, and placed consecutively in shallow drawers, with 50-100 in a drawer, while the classification may be made by means of a card catalogue, which would contain a numbered card for every individual, arranged in accordance with the system just described. I have arranged my own print collection upon this latter system, the prints being numbered and arranged consecutively in the order taken, while the corresponding cards are classified in accordance with the formulæ, those of the left hands taken first, and subdivided by those of the right.
As for the space required for a collection. If the prints are filed in shallow drawers or slides holding 50 sets each, a collection of 100,000 sets could be accommodated in 2,000 drawers; and allowing a frontage of 16 x 3 inches for each drawer with its surroundings, these would make a cabinet 6 feet high and composed of 56 perpendicular stacks which together would occupy a wall length of 75 feet. An 18-drawer card index with a capacity of more than 20,000 3 x 5-inch cards, as taken from a recent catalogue of office furniture, is 44 inches wide and 14 in height, and the five necessary to accompany the collection in question would form a single stack 6 feet high and 44 inches wide, which will add approximately four feet to the seventy-five given above. Thus a room having 80 feet of wall, linear measure, or a smaller one with a double stack running through the center, would be amply sufficient for the entire collection, properly arranged.
This calculation appears to answer in the affirmative the question of the practicality of keeping palm and sole records of all citizens as advocated in my previous article. M. Bertillon has pointed out the numerous cases in civil life in which one's identity is in peril, and looks forward to a future in which some record, based upon physical characters, will be made of every citizen, but the trouble and inconvenience attending his own system of measurements and the fact that they are applicable during adult life alone, would leave them hardly to be considered for such a purpose.
The palm and sole system, which I originally presented as an extension of the system of Mr. Galton, appears to supply the need in this respect, as the records are easily taken, unchanged from birth to death, quickly compared, either with the hands and feet themselves or with other prints, and capable of brief characterization and of accurate classification by means of simple formulae, to all of which may be added, as their most important advantage, that of absolute certainty, while the Bertillon measurements afford no more than a strong probability.
Prints could be taken in each township or municipality, and filed away in any convenient spot, perhaps the court house of each county seat. They could be taken in connection with the school registration, either when first entering, or better yet at the age of twelve to fifteen. There can be no question as to legal right in compelling such records, since there is no serious objection to compulsory vaccination, a far more serious operation, and one incurring a slight indisposition and a permanent change in the system, the nature of which is as yet unknown.
Similar records could be taken by the various civil and religious institutions in which the identity of an individual is apt to be called in question. Banks could require an imprint of the left palm upon the inside cover of bank-books; business men could issue checks with a fac-simile engraving of the palm of their own left hand covering the face; insurance companies could keep a palm and sole list of their clients; the Geary law would be rendered a certainty if the certificate issued to each Chinaman bore, besides the photograph, a single palm print, and churches could file away palm and sole prints with their baptismal records.
In the words of Bertillon, the founder of anthropometric identification: "La constitution de la personnalité physique et de l'indéniable identité des individus arrivés à l'âge adulte répond, dans la société moderne, aux besoins les plus réels, aux services les plus variés. . . . En un mot, fixer la personnalité humaine, donner à chaque être humain une identité, une individualité certaine, durable, invariable, toujours reconnaissable et facilement démontrable, tel semble l'objet le plus large de la méthode nouvelle. "[10]
- ↑ Bertillon, A. 'Instructions Signalétiques,' 1893. Introduction.
- ↑ A report of these investigations will shortly be published. See note, p. 396.
- ↑ See Am. Journal of Anat., Vol. 1, No. 4, November, 1902.
- ↑ 'Finger Prints,' Macmillan. 1892, pp. 107-108.
- ↑ "Ainsi la solution du problème de l'identification judiciaire consistait moins dans la recherche de nouveaux éléments caracteristiques de l'individualité que dans la decouverte d'un moyen de classification. Certes, je ne conteste pas, pour no parler que du procédé chinois, que les arabesques filigranées que presente l'repiderme de la face anterieure du pouce ne soiont à la fois fixes chez le même sujet et extraordinairement variables d'un sujet à un autre; et que chaque individu ne possède là une espèce de sceau original et bien personnel. Malheuresement il est tout aussi indéniable, malgré les recherches ingenieuses poursuivies par M. Francis Galton, en Angleterre, que ces dessins ne presentent pas par eux-même des éléments de variabilité assez tranchés pour servir de 0 base à un repertoire de plusieurs centaines de mille cas." Bertillon. 'Instructions Signalétiques,' 1893, Introduction.
- ↑ M. Bertillon's reply to the sending of my first paper is as follows:
Paris, le 12 Janvier 1903.
Monsieur.
J'ai pris grand interêt il la lecture de votre étude sur les lignes papillaires du pied et de la main, et je vous prie de recevoir tons mes remerciments pour l'obligeance que vous avez eue à m'adresser cette publication.
Conformement à votre désir, je vous transmets un examplaire de l'Introduction de l'ouvrage qui j'ai fait parêitre en 1893 sous le titre de "Instructions Signalétiques." l'editeur en est Mon. Durand, Rue Oberkampf No. 80 a Paris.
Veuillez agrier, Monsieur, l'expression de mes sentiments les plus distingués.
Le Chef du Service de l'IdentificationA. Bertillon.Monsieur H. Wilder a Northampton. - ↑ In Indian Service 1853-1878.
- ↑ This and other morphological points of which I shall make use in this article are from an unpublished paper upon the morphology of the subject, by an associate in my department, Miss Inez L. Whipple. At my suggestion Miss Whipple has undertaken the comparison of the human conditions of palm and sole with those of the lower primates and other mammals, and has studied also the ontogenetic development of the parts in man and other forms. This work, which is of great value in the present connection, will be published in full in a short time.
- ↑ Consult in this connection my comparison of identical twins, which are in all probability nearer alike in the palm and sole markings than are any other two human beings, and which nevertheless differ to a noticeable degree in the minutæ. Amer. Jour. of Anat., Vol. I., No. 4. 1902.
- ↑ 'Instructions signalétiques,' 1893, Introd., p. Ixxxiii.