The original plan was to limit the investigation to those babies born in the calendar year selected whose births had been registered, the purpose being to secure facts concerning a definite group and not to measure the completeness of birth registration. Shortly after beginning the work, however, agents of this bureau were told that the Serbian women in Johnstown seldom had either a midwife or a physician at childbirth; that they called in a neighbor or depended upon their husbands for help at such times, or that they managed alone for themselves, and that therefore their babies usually escaped registration. The omission of these babies meant the exclusion of a number of mothers in a group that was too important racially to be omitted from an investigation embracing all races and classes. Accordingly a list of babies christened in the Serbian Church and born in the year 1911 was secured and an attempt made to locate them. In addition an agent called at each house in the principal Serbian quarter to inquire concerning births in 1911. A number of unregistered babies of Serbian mothers were thus found and included in the investigation. The agents were sometimes approached by mothers of babies born in 1911 who resented being omitted from the investigation simply for the reason that their babies’ births had not been registered. The agents were therefore instructed to interview mothers thus accidentally encountered and to include their babies in the investigation. But no additional baptismal records were copied nor was a house-to-house canvass made of the city; in fact, no further means were resorted to to locate unregistered babies for the purpose of including them in the investigation. There were 1,763 certificates copied at Harrisburg, and 1,383 of the babies named in them were reached by the agents. In addition, 168 babies for whom there were no birth certificates, but who were located in the ways just noted, were included, making a total of 1,551 completed schedules secured. Of the 380 not included in the investigation there were 149 who could not be located at all; 220 others had moved out of reach—that is, into another city or State; 6 of the mothers had died; 3 could not be found at home after several calls, and 2 refused to be interviewed. From the following summary of data recorded on the certificates of the 380 unlocated babies just referred to it appears that the infant mortality rate (134.3) among them is almost the same as that (134) shown in Table 1 for babies included in the investigation. In reality, however, it is perhaps a little higher, as some of these babies no doubt died outside of Johnstown and their deaths were recorded elsewhere.