measured by a coefficient of association, the connection between intellectual weakness in the relatives and delinquency in the children rises to .36, a figure higher even than that obtained in the case of moral weakness. As with moral infirmity, so with intellectual: in none but a few exceptional examples is the association thus measured attributable exclusively to true biological inheritance; the faults and limitations of the father or the mother operate quite as much through the social environment they engender; and the dull parent, like the depraved, inevitably creates a home in which discipline is weak, and in which delinquency is only too readily fostered. The two remaining groups of inheritable conditions—namely, physical infirmities, and temperamental defects involving pathological symptoms-show comparatively small coefficients1.

The problem of the born criminal, however, is not answered and ended by a survey merely of hereditary conditions. What is inherited is necessarily inborn; but what is inborn has not necessarily been inherited. It is quite conceivable that a child may be afflicted with congenital propensities, without those propensities having first emerged in the previous history of his family.

To gain light upon this broader problem, I have endeavoured to separate all cases in which the factors, whether principal or accessory, were of a congenital type. Under this rubric I have included all such physical conditions as appeared to be directly inherited, all intellectual conditions that are now generally assumed to be innate (as mental deficiency and general dulness), all instances of general emotionality not due to adolescence, and all instances of a natural over-development of a primary instinctive disposition; examples of mere educational disability, instances of repressed complexes and of defective or undesirable interests, I have excluded. We have thus a division of cases and causes into those predominantly congenital and those predominantly acquired. I must insist once more that such a distinction is, and can be, nothing but a rough and somewhat speculative one.

Altogether, congenital factors, whether major or minor, are found

1 The four coefficients cited in the text are calculated from the frequency of the specified conditions among all the nearer relatives of the children. If we base the coefficients upon the appearance of the conditions among the parents only, the association is somewhat higher (see second column, Table IV). It will be noted that for physical conditions the coefficient is nearly doubled when we restrict our data solely to the parents. This, however, but confirms the suspicion that the working of such a factor is not entirely through direct biological inheritance. In many of these cases the parental condition, though often showing itself as some hereditary weakness in the child, also conduced to poor discipline in the home.

some 259 times per cent. among the delinquents, but only 92 times per cent. among the non-delinquents. Non-congenital factors are entered 208 times per cent. among the non-delinquents and among the delinquents 746 times, or, if we omit the psycho-analytic complexes (which were uninvestigated in the control group), 593 times. Thus congenital factors have been recorded among delinquents 2-8 times as often as among non-delinquents; and non-congenital factors 3-8 times as often, or, omitting complexes, 3.6. The major factor proves to be of a congenital type among 35.7 per cent. of the boys and among 40.9 per cent. of the girls. Thus, in well over one-third of the cases, but in rather less than one-half, some congenital weakness is the preponderant factor.

The share of congenital conditions in the production of delinquency is thus undoubtedly considerable. These are, indeed, the cases that are likely to be the most obstinate, and to be most in need of curative rather than punitive measures. But it would be a gross misconstruction-a mistake too commonly deduced from current fatalistic theories--to depict them as hopeless victims of their inborn nature. And at the same time it will be perceived that there still remains a large balance of delinquents -between 50 and 60 per cent. of the total-who owe their delinquency predominantly to the difficulties of their environment or to the events of their own past life.

The distribution of principal causes into congenital and non-congenital reveals proportions very dissimilar in my own cases to that announced by previous investigators. Among Healy's recidivists1, environmental influences formed the cardinal factor in only 25 per cent. of the cases and a subsidiary factor in as many as 76 per cent.; on the other hand, mental abnormalities and peculiarities, for the most part presumably innate, played a major part in 55 per cent. and a minor part in only 16 per cent. From these proportions we might be induced to argue that the prime causes of delinquency were, in the majority of instances, inborn psychological characteristics; and that the environmental factors were, as a rule, only accessory. Healy, however, has taken into account chiefly the more extreme psychological conditions-as feeblemindedness, hysteria, "epileptic mentality," markedly neurotic and psychotic states; and it would seem that the cases submitted to him, partly, no doubt, because they consisted of older, more hardened, and more habitual types, and partly perhaps because they were singled out by the court as in special need of examination at a psychopathic institute, comprised a number disproportionately large of gross aberrations of this kind.

Gruhle, in a small enquiry often cited, has attempted, on lines similar to the foregoing, to divide his cases into congenital and environmental types. The figures given by him, based upon the histories of 105 youths, strongly favour the importance of inborn constitution. His proportions are the following: congenital cases 40.9 per cent., environmental cases 9.1 per cent., congenital and environmental factors being 1 Loc. cit. pp. 130–132.

J. of Psych. (Med. Sect.) III

2

equally balanced in 40.9 per cent.1. Lund's data, on the other hand, emphasise rather the power of environment; his figures are as follows: in one investigation, the congenital cases numbered 23.4 per cent., and the environmental 65-7 per cent., environmental and congenital factors being equally balanced in 10.9 per cent.; in a second series, the congenital cases numbered 21.1 per cent., the environmental 64.9 per cent., the factors being again equally balanced in 11·9 per cent.2.

In one significant feature this and most foreign studies of recent date seem broadly in agreement. Whatever may be our various theories upon the relative importance of the two factors, most cases of juvenile delinquency are, after all, admitted almost universally to be the joint product of social environment and congenital constitution, working together and playing the one upon the other. Gruhle, for example, so often cited in favour of the inborn factor, nevertheless recognises a "mixed type" of criminal, neither purely congenital, nor purely environmental, comprising practically 70 per cent. of his cases. Lund makes a similar concession, though to a less generous extent; in one of his investigations 36·0 per cent., and in another 39.8 per cent. belonged to this mixed type. Both writers, it is true, claim, even within this blended group, to discover now environment, and now constitution preponderating; but, as just remarked, they still find an appreciable proportion in which the two tributaries are so equally balanced that not even this distinction can be hazarded.

I personally should hesitate to recognise any "pure types" whatever. The practical test of a pure congenital type, is, I suppose, that, no matter how perfect the environment found for them, in the most comfortable of homes, in the best disciplined of institutions, they still evince propensities to crime. By this criterion 18 per cent. of my cases show hopeless and ineradicable delinquency. Their irremediable condition, however, I incline to attribute less to original endowment than to long-standing habit. They are, nearly all of them, old offenders. But I suspect that, detected and treated at an early stage, and provided with a suitably adjusted environment, were it only the environment of a segregated colony, the worst and dullest of them might yet have been rescued, and converted into quiet workers and law-abiding members of their own little community.

II. ENVIRONMENTAL FACTORS.

In the child's environment the adverse conditions noticed have been grouped under five main heads (Table II). Of these, defective home discipline seems the one most vitally connected with juvenile crime. Between this factor and delinquency the average coefficient of association (Table IV) is 46. A vicious home atmosphere is almost equally detrimental. Defective family relationships, and undesirable surroundings outside the home, exert a smaller, yet still an appreciable, effect. The pressure of poverty, with its more immediate concomitants, seems, at any rate in its direct relation to delinquency, of comparatively slight significance.

1 Loc. cit. p. 23. The cases here classed together as congenital he has sub-divided into a purely congenital and a predominantly congenital group; and similarly with the environ. mental.

2 Loc. cit. p. 233.

A glance at the particular features classed under these five headings (Table VI) reveals still further differences within each broad category.

TABLE VI. II. Environmental Conditions.

The classification used is that suggested by Charles Booth (Life and Labour in London, Vol. I. Pt. I, pp. 33 et seq.). His categories are as follows:

A. Occasional labourers; loafers; street-sellers; the criminal and semi-criminal. (In my cases, 7-3 % of the boys, and 5-4 % of the girls came from homes of this type.)

B. Irregular earnings; casual labour.

C. Intermittent earnings; seasonal labour.

D. Small regular earnings.

E. Regular standard earnings; artisans; small shop-keepers (with no assistants).

F. High class labour, well-paid; foremen; best-paid artisans.

G. Lower middle class; shop-keepers; tradesmen; small employers; clerks.

H. Upper middle and upper classes; servant-keeping class.

The grouping of the categories in pairs, and the titles for the composite categories, are Booth's. Since the non-delinquent children were definitely selected as belonging in the same proportions to the same social classes, the percentages would be the same as for the delinquents. In the last column, therefore, I have given instead Charles Booth's estimates for the relative proportions of the whole population of London comprised in the several groups. The borough in which my survey was made shows, according to his tables, almost exactly the same proportions as London taken as a whole. 2 Tenements with more than two occupants per room.

3 Chiefly, absent for a long period upon military service; in some instances, recently returned. Also includes illegitimate cases where the father is not living with the mother.

Includes also the few cases born out of wedlock where the mother and father afterwards married, or lived together as married.

* Figures in brackets are not included in the totals. The discrepancy between the totals for the boys and girls and the average total is accounted for in footnote (1).

8 Chiefly parents (or foster parents) living together unmarried.

8, 9, 10 In this section only present or recent misconduct in the child's own home is included: the figures given under "hereditary conditions" include also offences committed in the remote past or away from the child's own home.

10 Chiefly stealing or habitual gambling and betting.