That personage on horseback dominating the piazza of the Capitol, were he alive and not a bronze statue, would I think be very happy about this Symposium now opening.

I may even imagine that Marcus Aurelius would leave his role of sceptical emperor-philosopher to assume that of paterfamilias, smiling benignly upon us all, as he was the father of twins: Commodus, emperor after him, and Antonius Geminus.

He had a special medal struck, when these twins were born to him by Faustina, bearing her image and that of Juno Lucina, goddess of fertility. He then had six children, and Faustina was to bear him seven more — quite a large family, as is often the case when there are twins.

Yes, I am sure that Marcus Aurelius would rejoice at the opening today in Rome of this first Symposium of twin researchers, their object being not only to know twin better, but also to assist them, and because the Symposium is the first, the beginning, we hope, of a series of many more most fruitful ones.

This report is concerned with the twin births which occurred in the Collaborative Study of Cerebral Palsy, Mental Retardation and Other Neurological and Sensory Disorders of Infancy and Childhood. This is a collaborative endeavor of 14 institutions throughout the US and the National Institute of Neurological Diseases and Stroke, to observe and study factors which affect the parents before a child is born and to relate them to the outcome of pregnancy. To this end, we are following approximately 60 000 pregnant women from the first months of their pregnancy, through labor and delivery and up to the seventh year of the study child's life. The Study population is about 46% White, 46% Negro, 7% Puerto Rican, and the rest a variety of other ethnic groups. The prospective nature of the study makes possible the recording of observations with high accuracy and the avoidance of biases inherent in other types of investigations. These advantages should also be reflected in the collection and evaluation of twin data.

Nella consulenza eugenica, il metodo formale, fondato sull'esame dei rapporti mendeliani attesi in funzione del tipo di trasmissione del carattere, dovrebbe essere la norma, ma le limitate conoscenze di genetica umana e medica lo rendono inapplicabile per la maggior parte dei casi.

È necessario, quindi, ricorrere al pronostico empirico per poter soddisfare alle domande che vengono rivolte al consulente eugenico. Tra le domande che più spesso ricorrono vi è quella posta dalle madri di gemelli per conoscere quale probabilità esse hanno di un novello evento gemellare nei loro concepimenti.

Per dare una risposta a questo interrogativo e chiarire contemporaneamente il quesito che, anche se più raramente, viene posto da madri, esse Stesse gemelle, su quale sia la probabilità di avere figli gemelli, abbiamo predisposto due diversi rilevamenti tra le oltre 14000 schede della gemelloteca dell'Istituto Mendel di Roma.

Il primo rilevamento, limitato a 2000 schede contenenti alberi genealogici con dati certi, ha dato come risultato un numero di genitori di gemelli, anch'essi gemelli, di 58; cioè, di 58 su 4000 genitori complessivi.

The rate of twinning, both MZ and DZ, as determined by the Weinberg formula, remained fairly static in Scotland up to 1958, when it fell (Fig. 1), due to a fall in the DZ twinning rate. These figures are based on the returns from the Registrar General for Scotland.

As age has a marked influence on the incidence of twinning, it was thought that the difference might be due to a reduced number of older women having babies. However, this does not seem to be the whole answer (Fig. 2). The most marked change would be expected in the women over 35, but this did not occur. It was not possible to look at other factors such as parity, social class and height in the figures for Scotland. The details of all births in the City of Aberdeen are known because of the very high hospital delivery rate, and more detailed studies have been made on the twinning rates for Aberdeen.

These figures, again, showed that from 1959 onwards there was a fall in the twinning rate in Aberdeen and this was due to the fall in the number of DZ twins. The fall in incidence occurred both in Primigravidae and in multigravidae (Tab. I), and more in older than in younger Primigravidae (Tab. II), but there was little difference by age in the fall in multigravidae (Tab. III).

Another factor which is related to the incidence of twinning is the height of the mother. This, in turn, reflects the social class, as tall women are more commonly from the upper, and small women from the lower social classes. The rate of twinning in Aberdeen was highest in the tall women before 1954 (Tab. IV). The rates in all women fell from 1954 onwards, but most dramatically in the tall women. The rates from 1959 onwards show very little difference according to height.

The probability of a twin birth depends strongly and almost equally on the age of the mother and on the number of children she has had. Even when these variables are held constant, the frequency of twinning varies between countries or geographic regions and between years or decades. This variation has been thought perhaps to result from economic and environmental factors, but another explanation can be suggested.

Bulmer (1959) found that twins are frequent among the earliest conceptions after marriage. Eriksson and Fellman (1967) found twins to be frequent among illegitimate births and suggested that the mothers of twins conceive more easily than other women; in other words, that they are more fecundable. We can present some new evidence for this hypothesis, and we believe that high fecundability in the mothers of twins can account for variations in the twinning rate.

Fig. 1 shows the wide range of twinning rates in the US white population, in different parts of the country and for three periods, centering on 1937, 1954 and 1964. The nine geographic divisions are shown in descending order according to the twinning rates in 1937. The two regions at the right, with the lowest rates in 1937, had nearly stable twinning rates over the three time periods. They are the oldest and most urban parts of the country.

In Bulgaria, in un periodo di 88 anni, sono nati circa 12 500 000 bambini, di cui circa 300 000 gemelli, 4000 trigemini e 100 quadrigemini. Si è dunque avuta una frequenza di 1/80 per i parti gemellari, 1/6 000 per quelli trigemini, e 1/500 000 per quelli quadrigemini. In 69 anni (1894-1962) si sono verificati 23 parti quadrigemini, cioè, uno ogni tre anni.

Fino al 1912, la frequenza dei parti gemellari è stata dell'1.5%; 50 anni più tardi (1962), tale frequenza diminuisce della metà. Dal 1904 al 1912, su una popolazione di 60 00 000 di abitanti, sono nati 168 000-186 000 bambini all'anno; dal 1954 al 1962, su una popolazione di 8 000 000 di abitanti, ne sono nati 135 000-150 000.

La brusca diminuzione della natalità gemellare nel paese è più indicativa se si fa il confronto su 10 000 nati. Per i maschi, la frequenza cade da 221/10 000 a 165/10 000; per le femmine, da 234/10 000 a 170/10 000.

Twinning is regarded as a very important phenomenon among the Yorubas in Western Nigeria. In this population, special ceremonies are performed to mark twin deliveries, and twin babies and their parents are given special names. Twins are therefore more readily identified, and data relating to twinning in families are likely to be fairly accurate.

Furthermore, the practice of polygamy in Western Nigeria provides a good opportunity for studying both maternal and paternal factors in the inheritance of twinning. The high incidence of twinning in the population (45‰: Nylander, 1969) is also an additional advantage, since it allows a large number of twin maternities to be studied in a short space of time.

Among 18 737 deliveries in the three major hospitals in Ibadan, Western Nigeria, between October 1967 and June 1969, there were 977 twin maternities. Every mother was interrogated immediately after delivery to find out whether she or her husband was a twin or had twins previously (including those delivered by the husband's other wives), and whether she or her husband had twin siblings. Obstetric data, including maternal age and parity, were collected for each patient, and zygosity of the newborn twins was determined by sex, placentation, blood groups, placental enzymes, haemoglobin types, and G6PD electro-phoretic studies, as will be described in a subsequent paper.

Nelle ricerche sulla gemellazione la madre dei gemelli è la candidata, ma il carattere ereditario si manifesta nella generazione successiva.

Partendo da considerazioni probabilmente diverse, molto studio è stato dedicato direttamente alla candidata, ai suoi caratteri fisici e fisiologici. La generazione successiva è stata invece alquanto trascurata, sicché, quando abbiamo voluto considerare l'andamento del peso alla nascita dei fratelli di gemelli, non ci è stato possibile rin-tracciare precedenti bibliografici.

Il peso alia nascita è sotto l'influenza di innumerevoli fattori ed ovviamente non è un carattere ereditario. Ma, se la gemellazione è provocata da un sistema ormonico controllato dal genotipo, è probabile che questo sistema agisca anche sul peso alla nascita dei fratelli di gemelli.

Frequency and inheritance of twinning have been studied from urban and rural parts of Madhya Pradesh. Family data conform to previous findings by other authors, for other countries, that DZ twinning is influenced by the mother's genotype.

Of special interest has been a rural family in which a female, herself a cotwin of DZ pair, married twice, each time giving birth to DZ twins.

The high rate of twinning in rural areas is suggested to be due to higher coefficient of inbreeding, better physical condition of mothers on the countryside, higher mean birth order, and higher mean age of mothers. Lack of family planning may have also played a role.

Nos explications ont pour but de faciliter la compréhension de la concordance et surtout la discordance des jumeaux MZ et DZ moyennant les analyses chromosomiques. Dans les dernières années tout un nombre d'études concernant les jumeaux ont été publiées où chaque jumeau accusait le même phénotype et caryotype. Il s'agit de jumeaux MZ avec les syndromes de Down (Mikkelsen et Melchior, 1962), de Klinefelter (Bertrand et al, 1966; Holub et al, 1958; Nowakowski et al, 1963; Vague et al, 1968) et de Turner (Decourt et al. 1964; Frasier et al, 1961; Nance et Uchida, 1964; Turner et Zanartu, 1962); aneuploïdies doubles (Hustinx et al, 1961; Turpin et al, 1964); mosaïques gonosomiques (Mikkelsen et al, 1963) et caryotype XYY (Rainer et al, 1969).

Le problème de la discordance des jumeaux MZ nous préoccupe davantage que celui de la concordance, car en effet la discordance a été l'objet de maintes études. Le résultat n'est pas satisfaisant pour tous les jumeaux étudiés (Nielsen, 1967; v. Verschuer, 1927). À ce sujet, je tiens à rappeler la discussion qui a eu lieu dans l'Institut Mendel, au cours du Premier Symposium International de Génétique Médicale du 6 au 7 septembre 1953, entre Kallmann et v. Verschuer. Parmi les premiers savants il faudrait citer Meirowsky (1926) qui expliquait les causes de la discordance par un partage inégal du plasma lors du clivage du disque embryonnaire.

The incidence of twinning in Western Nigeria is very high, being 45‰ (Nylander, 1969a), as compared with 11-12‰ in Caucasian populations (Registrar General, 1967). A recent study of twinning in Ibadan, in which zygosity of individual twin pairs was determined, has shown that the pattern of placentation and zygosity in this population is also very different from that in Caucasian populations.

As many as 1475 consecutive newborn twin pairs delivered in three hospitals in Ibadan between March 1967 and April 1969 were investigated.

The placentae of the twins were examined macroscopically by the author personally to determine the form (whether single, fused, double or separate) and the membrane relationships (whether monochorionic or dichorionic) as shown in Figs. 1 and 2. The membrane relationship was confirmed by histological examination.

Cord blood samples were collected from each twin pair and studied for (1) ABO, Rh (using antisera, C, c, D, E, e), MNSs and Gonzales blood groups and (2) G6PD electrophoretic pattern. Other blood groups like Fya, k and Jk, used initially, were discontinued later, because the gene frequencies in this population are such that they are uninformative for zygosity determination.

Samples of the placenta of each twin were sent in refrigerated containers to the Galton Laboratory, London University, for placental enzyme studies — phosphoglucomutase and peptidase (Hopkinson and Harris, 1968; Lewis and Harris, 1967).

De nombreux auteurs estiment, sur la base de ressemblances électroencéphalo graphiques observées au sein d'une même famille ou chez des paires de jumeaux MZ, que l'EEG est déterminé génétiquement (Lennox et al, 1945; Weiss et Rudkowka, 1962; Vogel, 1965; Heuschert, 1965). Beaussart et al (1959) pensent même que cet examen peut servir de critère pour le diagnostic de monozygotie.

Nous avons tenté de vérifier si l'étude de la maturation de l'EEG conduisait aux même conclusions. Nous avons comparé le degré de maturation de l'électrogenèse corticale che: des paires de jumeaux concordants et discordants pour les groupes ABO, MN et Rh. Nous nous sommes adressés à des enfants et des adolescents. On sait que, chez ceux-ci, l'EEG comporte, en plus du rythme α, des ondes θ dont les fréquences sont de 4-7/sec, et des ondes δ de fréquences inférieures à 4/sec. Ces ondes lentes diminuent progressivement avec l'âge poui disparaître chez l'adulte normal. À un âge donné, leur nombre est très variable d'un suje à l'autre. Le degré de maturation de l'EEG peut donc être apprécié de façon simple par l'évaluation de la proportion d'ondes lentes du tracé.

Notre étude a porté, d'une part, sur 24 paires de jumeaux concordants sanguins, âgés en moyenne, de 10.5 ans (6-18 ans) et, d'autre part, sur 34 paires de jumeaux discordants âgés, en moyenne, de 9.3 ans (6-16 ans). Les sujets étaient d'origine rurale et appartenaien à un même groupe socio-économique. Le quotient intellectuel moyen était de 92.8 pour les concordants et de 93.5 pour les discordants.

Dans le laboratoire d'anthropologie de l'Institut Royal des Sciences Naturelles de Belgique, le Prof. Twiesselmann a organisé, en 1961, une enquête gémellaire au cours de laquelle une centaine de paires (jumeaux de même sexe) et de couples (jumeaux de sexes opposés) furent analysées. Seules 63 paires pour lesquelles nous possédons des données biométriques complètes seront envisagées dans cette note.

Neuf sérums sanguins — respectivement deux, deux et cinq pour les systèmes ABO, MN et Rh — permirent de séparer les 63 paires en deux groupes:

1) 30 paires de jumeaux concordants pour les neuf sérums;

2) 33 paires de jumeaux discordants pour au moins un sérum.

Les 33 paires étant certainement DZ, nous nous proposons d'évaluer, en termes de probabilité, le nombre k de paires DZ qui peuvent se cacher parmi les 30 paires de jumeaux concordants, dont la majorité est MZ.

Pour chacune des 30 paires de concordants, Mme Vrydagh a calculé la probabilité de monozygotie en utilisant la méthode de Maynard-Smith et Penrose (1954) qui est basée sur le théorème de Bayes de la probabilité des causes. Pour nos 30 paires de concordants, ces probabilités vont de 83 à 92%; elles auraient été plus élevées si plus de sérums sanguins avaient été utilisés.

Although published reports of cancer in twins are not numerous, there are case reports of concordance in a variety of tumors in twins. Cancer of the breast is most frequent; concordance has also been noted in uterus, gonads, eyes, stomach and rectum. In general, these reports tend to support the theory that genetic factors operate either in the concordance of cancer or in the site of the specific tumors. The significance of concordance remains in doubt, however. Reported study groups are not comparable because the materials and methods are disparate. In many reviews the methods of zygosity determination are questionable. The fact that not all cases of cancer in twins are reported introduces a constant source of bias. The following review illustrates this well.

In 1940, Madge T. Macklin reviewed the publications of tumors in MZ and DZ twins and concluded that tumors affected MZ twins far more than both members of a DZ pair. She observed concordance in tumor type, site, and age of onset, as more frequent in MZ than DZ twins. She later (1947) reexamined the same material and some additional twin pairs and concluded that MZ twins have identical tumors far more frequently than DZ twins. She emphasized that rare types of tumors were genetically determined.

In 1948, Busk et al reported on a series of 185 twin study pairs from the Danish Cancer Registry; they concluded that (1) there was a tendency toward a higher incidence of cancer in partners of MZ than of DZ cancerous twins; (2) these deviations from expected values were not considered statistically significant; and (3) there was a tendency for tumors in MZ twins to affect corresponding organs in both partners, this being not the case in DZ twins.

Studies on cancer in man and pure strain animals have shown that the malignant neoplasm, both in animals and in human beings, is not a single, genetically homogeneous disease: it consists of a lot of forms differing in localization, histological structure and etiopathogenesis. This circumstance is not always taken into account by the investigators in cancer research by the twin method, and especially in the determination of within-pair concordance/discordance, which is often a source of misunderstandings and faulty conclusions.

In order to establish a more objective criterion for estimating the real twin concordance in the studies of such diseases as cancer, we have worked out a quantitative index called Concordance Index:

where: CI = Concordance Index %; a = number of twin pairs, MZ or DZ, which are fully concordant for the principal features; b = number of partially concordant twin pairs; n = total number of twin pairs, MZ or DZ.

The accuracy of twin zygosity determination acquires a particular importance in cases of low penetrance of the disease. In such cases, even a minor inaccuracy in the zygosity diagnosis may considerably influence the twins' CI, and always in the same direction: any mistake in the determination of twin zygosity always leads to an artificial diminution of the role of hereditary factors. Therefore, we consider that the questionnaire method of twin zygosity determination, as well as any other insufficiently objective method, should not be used in the genetic studies on cancer.

We undertook an extensive review of world-wide clinical reports of leukemia in twins. Our survey was a careful one; exclusions are inevitable, but, if so, are inadvertent, except for undocumented “personal communication” cases. These have sometimes been included in large reviews of either leukemia or leukemia in twins. Also excluded are cases mentioned only tangentially in reports of cancer in large population groups, if no clinical details were given. In some instances, we were able to adapt available clinical information to our tables.

Leukemia appears to occur at four periods in the lives of twins: (1) the perinatal-congenital period; (2) early childhood; (3) late childhood; and (4) adulthood. Those cases occurring around the time of birth and the first year of life are combined under the title: Perinatal-Congenital. In Tab. I, Early Childhood extends from about 2 to 7 years, and Late Childhood from 7 through 12 years. All other cases have been considered as Adult. Two cases occurring in teenagers have been placed in this group in accordance with clinical practice. We have compiled 62 cases of leukemia in twins from clinical reports from the world publications in the following categories: Perinatal-Congenital, 15; Early Childhood, 26; Late Childhood, 9; Adulthood, 12. Comparative clinical details of all these cases will be published in a subsequent report.

Except for those instances about which writers admitted that the diagnosis of zygosity was in question, most reported clinical cases occurred in MZ twins. There were 39 cases of leukemia in MZ twins, 13 cases in DZ twins and 10 cases in whom zygosity was in doubt.