1 The best overview of the nature of European infant abandonment in this earlier period is Boswell, John, The kindness of strangers (New York, 1988).Google Scholar

2 See Kertzer, David I., ‘Child abandonment in European history’, Journal of Family History 17 (1992), 13–19Google Scholar, and Ransel, David, ‘Child abandonment in European history: a symposium’, Journal of Family History 17 (1992), 19–23.Google Scholar

3 Osnaghi, Luisa Dodi, ‘Ruota e infanzia abbandonata a Milano nella prima meta dell' ottocento’, in Politi, G., Rosa, M. and Peruta, F. Delia eds., Timore e carità. I poveri neiritalia moderna (Cremona, 1982), 428.Google Scholar

4 Trisciuzzi, Leonardo and De Rosa, Diana (I bambini di Sua Maesta: esposti e orfani nella Trieste del 700 (Milan, 1986), 75)Google Scholar argue that before the nineteenth century the main concern of the authorities in Italy about foundlings' deaths lay in ensuring that the wetnurses did not continue to receive payments, along with concern for the cost of funerals. With the new century, ‘the death of the child becomes the central moment in the functioning of the institution and how to prevent it becomes an object for reflection’.

5 Kertzer, David I., Sacrificed for honor: Italian infant abandonment and the politics of reproductive control (Boston, 1993).Google Scholar

6 An excellent overview of the European situation, with figures for levels of infant abandonment, is provided by Volker Hunecke. ‘Intensità e fluttuazioni degli abbandoni dal XV al XIX secolo’, in Enfance abandonnée et société en Europe XIXe–XXe siecle (Rome, 1991), 27–72.Google Scholar Other chapters in that volume detail the systems of infant abandonment found in France, Belgium, Russia, Spain, Portugal and Austria. For recent sources in English, see, among others: for Portugal, Isabel Dos Guimaraes Sá, The circulation of children in eighteenth-century Southern Europe: the case of the foundling hospital of Porto (unpublished D.Phil, thesis, European University Institute, Florence, 1992)Google Scholar; for France, Fuchs, Rachel, Abandoned children: foundlings and child welfare in nineteenth-century France (Albany, 1984)Google Scholar; for Russia, Ransel, David, Mothers of misery: child abandonment in Russia (Princeton, 1988)Google Scholar; for Spain, Sherwood, Joan, Poverty in eighteenth-century Spain: the women and children of the Inclusa (Toronto, 1988)Google Scholar; for Ireland, Robins, Joseph, The lost children: a study of charity children in Ireland, 1700–1900 (Dublin, 1989).Google Scholar

7 The difference between the two systems was dramatically illustrated by Tocci, Guglielmo (Gli esposti e l' organizzazione della carità pubblica della Provincia di Cosenza (Bari, 1878), 58)Google Scholar, who tells that of the 1,200 babies abandoned every year in the southern province of Cosenza in the 1870s, half were sent to the Cosenza foundling home and half were sent directly by rural mayors to local wetnurses. Of those sent to the foundling home, he claimed, 90 per cent died in their first year, compared to just 25 per cent of those sent directly to wetnurses.

8 Da Molin, Giovanna, L' infanzia abbandonata in Italia nell' etá moderna (Bari, 1981), 109–10.Google Scholar On infant mortality in Naples, see also De Crescenzio, Nicola, I brefotrofi e la esposizione del bambini (Naples, 1873), 224–32Google Scholar, Mario, Jessie White, La miseria in Napoli (Florence, 1877), 91–2Google Scholar, and Stella, Pietro and Da Molin, Giovanna, ‘Offensiva rigoristica e comportamento demografico in Italia (1600–1860)’, Salesianum 40 (1978), 3–55.Google Scholar

9 For an example of the search for these qualities, see Bruscoli, G., Lo Spedale di Santa Maria degl'Innocenti (Florence, 1900), 207–8Google Scholar, on the Florentine ospizio.

10 The impact of measures taken later in the nineteenth century against the spread of syphilis in delaying foundling placement in Florence, with negative consequences for foundling survivorship, is described in Viazzo, P. P., Bortolotto, M. and Zanotto, A., ‘Child care, infant mortality and the impact of legislation: the case of Florence's foundling hospital, 1840–1940’, Continuity and Change 9 (2), 1994, 243–69.CrossRefGoogle Scholar

11 The movement to close the wheels and the foundling homes is discussed in Kertzer, , Sacrificed for honor, 154–69.Google Scholar

12 The existence of this extremely valuable archival source was first brought to my attention by the paper of Eugenio Sonnino, who has previously used this source to examine the topic of mortality among abandoned babies in Italy in this period. Sonnino, Eugenio, ‘Esposizione e mortalitá degli esposti nello Stato Pontificio agli inizi dell' Ottocento, secondo le statistiche raccolte da Leopoldo Armaroli’, in Enfance abandonnée et société en Europe XIVe–XIXe siècle. Actes du colloque (Rome: Ecole Francaise de Rome, 1991), 1065–6.Google Scholar

In calculating IMR for foundling homes, we try to exclude those babies who are left dead at the foundling homes. In some cases we must rely on statistics where it is unclear how such infants were treated.

13 Angeli, Aurora, ‘Caratteristiche, mortalità ed inserimento sociale degli esposti nell' ospedale di Imola nei secoli xviii–xix’ in Boutry, P., ed., Enfance abandonée, 123–49.Google Scholar

14 Onger, Sergio, L' infanzia negata: storia dell' assistenza agli abbandonati e indigenti a Brescia nell' ottocento (Brescia, 1985).Google Scholar

15 Tocci, Guglielmo, Gli esposti e l' organizzazione dellà carita pubblica della Provincia di Cosenza (Bari, 1878), 61.Google Scholar

16 Following Unification, Italy's crude mortality rate (deaths per 1,000 population) declined from 30.3 in 1862–1871 to 23.7 in 1892–1901 (Panta, Lorenzo Del, Evoluzione demografica e popolamento nell' Italia dell' Ottocento (Bologna, 1984), 21).Google Scholar

17 Panta, Del, Evoluzione demografica, 11–15.Google Scholar

18 Residori, Sonia, ‘Tra demografia storica e storia della popolazione. Una comunità, una regione: Lendinare e il Veneto nell' 800’, Annali Veneti, Società, Cultura, Istituzioni 1: 1 (1984), 59.Google Scholar

19 Bellettini, Athos and Samoggia, Alessandra, ‘Evolution differentielle et mouvement saisonnier de la mortalité infantile et enfantine dans la banlieue de Boulogne (XVIIe–XIXe siècles), Annales de Démographic Historique (1983), 208–10.Google Scholar

20 ibid., 200.

21 Panta, Del, Evoluzione demografica, 55.Google Scholar In the 1880s, Berti, Giovanni (‘Ricerca fatta sulla mortalità nel l anno di vita dei bambini allattati dalle madri nella campagna bolognese’, Bullettino delle Scienze Mediche di Bologna (1886), series VI, 17, 357–8)Google Scholar, one of the medical directors of the Bologna foundling home, surveyed 667 women who applied to become paid external wetnurses. He found an IMR of 257 for their 2,449 live births.

22 The fact that unwed mothers were forced to abandon their babies did not lead to total exculpation for their doing so. In the late eighteenth century, for example, an observer of the high mortality among foundlings in Modena concluded that a woman who produced an illegitimate child was guilty of behaviour that was ‘inhumane, indeed cruel’; see Ricci, Lodovico, Riforma degli istituti pii delta città di Modena (Modena, 1797), 203.Google Scholar

23 Raseri, Enrico, ‘I fanciulli illegittimi e gli esposti in Italia’, Archivio di Statistica 6: 1 (1881), 12.Google Scholar

24 This is discussed for the Italian case in Kertzer, , Sacrificed for honor.Google Scholar One finds the same observations made linking poor foundling health to unmarried women trying to abort themselves in France at the time, as by Nepveur, M. L., De la mortalité des enfants trouvés en France, et à Rouen en particulier (Rouen, 1851), 8.Google Scholar

25 Luè, Giambattistia, Esposti (Milan, 1907), 59–60.Google Scholar In France this phase is most closely identified with abortionists.

26 Sussman, George D., Selling mother's milk: the wet-nursing business in France, 1715–1914 (Urbana, 1982).Google Scholar

27 Baret, Jean-Pierre, Dufour, Corinne and Renard, Jacques, ‘La mort des enfants trouves, un drame en deux actes’, in IUSSP Conference on Child and Infant Mortality in the Past (Montreal, 1992) (hereafter IUSSP 1992), 4.Google Scholar

28 Griffini, Romolo, ‘Delia mortalità dei bambini: della profilassi della sifilide infantile e da allattamento’. Annali Universali di Medidna 206 (1868), 275.Google Scholar

29 The Bologna foundling home medical report for 1879–1882 records that while the mean number of infants in the foundling home ranged from 22 to 33 in these years, there were various times when as many as 60–70 nursing infants were present. Yet the mean number of internal wetnurses in these years remained around 12 (Bruers, Emmanuele and Berti, Giovanni, ‘Rendiconto statistico dello stabilimento esposti e maternita di Bologna, pel quadriennio 1879–1882’, Bullettino delle Scienze Mediche di Bologna, series 6, 16 (1885), 166).Google Scholar

30 Onger, , ‘L' infanzia negata’, 78.Google Scholar

31 Bruscoli, Gaetano, Lo spedale di Santa Maria degl' Innocenti (Florence, 1900), 153Google Scholar, quoting Dr Calosi's report. See also the article by Bartolotto, Viazzo and Zanotto, , ‘Child care, infant mortality and the impact of legislation’, this volume.Google Scholar

32 Mantica, Nicolò, L' ospizio provinciale degli esposti e delle partorienti di Udine (Udine, 1900), 191.Google Scholar

33 These effects are nicely described in the report of the Statistica medico, primo saggio di Statistica medico nell' imerno del brefotrofio romano, anno 1867–1868 (Rome, 1869), vi, regarding the Roman foundling home.Google Scholar

34 From the archives of the Archivio Storico Provinciale di Bologna, Arch. Esposti, 22 January 1866, ‘Risposta ai quesiti intorno agli esposti… della Prefettura di Bologna’.

35 Given the length of time covered by the Restoration, we looked at both an earlier (1829–1830) and later (1849–1850) stage. We combined these two couplets of years only after determining that the mortality patterns were similar in each.

36 These figures include those babies who arrived dead. These are not counted in the various analyses of mortality and time-to-wetnurse-placement below.

37 Bellettini, and Samoggia, , ‘Evolution differentielle’, 208–10Google Scholar, and Panta, Del, ‘Infant and child mortality’, 16.Google Scholar

38 Using these figures, we can say that the Bologna foundlings' infant mortality disadvantage declined from an IMR 121 per cent higher than that of the general population to one 45 per cent higher.

39 Unfortunately, we have no good information on this point, though it is possible that by our last period the availability of improved transportation might have led to fewer babies from far-flung parts of the province arriving at the foundling home in poor health.

40 It has been argued, however, that deaths due to congenital problems are mostly clustered in the first few days following birth, and thus exogenous factors become a significant cause of death well within the first month. The concept of endogenous infant mortality was introduced by Bourgeois-Pichat, J. in ‘La mesure de la mortalité infantile’, Population 3 (1951), 233–48.CrossRefGoogle Scholar For a recent discussion of the use of the concept, see Lalou, R., ‘La mortalité endogène en Nouvelle-France’, IUSSP 1992.Google Scholar

41 Panta, Del, ‘Infant and child mortality’, 16.Google Scholar

42 This analysis excludes those few children who were left at the foundling home at above the age of six days.

43 Another factor which might be considered is seasonality of birth in the general agricultural population. Insofar as births show seasonal differences, of course, the number of potential wetnurses could change from season to season. However, this should only have an effect on the problem before us if illegitimate births showed a different seasonal pattern from legitimate births. This is an area that needs to be explored further.

44 Breschi, Marco and Bacci, Massimo Livi, ‘Month of birth as a factor of children's survival’, in IUSSP 1992, 21.Google Scholar

45 On the early existence of the Bologna wheel, see Bianchi, Adanella, ‘La “famiglia’ dell' Ospedale tra XVI e XVIII secolo’, in Amministrazione Provinciale di Bologna, Assessorato alia Cultura, I bastardini: patrimonio e memoria di un ospedale bolognese (Bologna, 1990), 45.Google Scholar

46 The Cox model is quite extensively used in demographic, biomedical settings, and has spawned a large literature. It produces a measure of the degree to which each covariate shifts the underlying hazard of infant mortality up or down. Estimation of the Cox model yields coefficients that, once exponentiated, indicate the relative risk factor for that covariate. For example, for a coefficient estimate of 0.1, the associated relative risk is e^0.1 or e^1.105, indicating that the presence of this covariate is associated with a 10.5 per cent rise in the risk of mortality throughout the observation period.

47 We first separately estimated models for each of our three periods. We present this combined analysis for the sake of simplicity, but it should be recognized that pooling periods tends to average out any differences that develop in the effects of the covariates over the three periods.

48 This link has also been made for the Florentine foundling home in the late nineteenth and early twentieth centuries by Viazzo, Bortolotto and Zanotto (‘Child care, infant mortality and the impact of legislation’).

49 A discussion of these causes may be found in Kertzer, Sacrificed for honor. They include both a decline in the influence of the Church following Italian Unification and a fiscal crisis suffered by the new régime, which made the huge costs of the foundling system a tempting target.