The study was carried on through the four seasons of the year. We considered some of the environmental and other factors which may have an influence on con¬ tact infection, such as poor economic status of the family with accompanying overcrowded living condi¬ tions, seasonal variation, the age factor and the occur¬ rence of acute colds in the family.All the various types of pneumococci (excluding types I and II) were as prevalent in the general popu¬ lation as in the pneumonia contact group. Types III, VI and XVIII were encountered most frequently.Types I and II, though responsible for more than half the cases of lobar pneumonia, were rarely encoun¬ tered in the general population ; but type II was three times more prevalent and type I six times more preva¬ lent in the contact group than in the control group. This observation is in accord with the observations of other workers in this field.Correlation tables of the prevalence of the specific homologous types of pneumococci in a case of lobar pneumonia and its family contacts indicate that the only strains that had a definite epidemiologie signifi¬ cance were types I and II. A fairly high correlation occurred with types V and VIII, but the numbers of persons exposed were small-thirty-two persons to type V and thirty-one persons to type VIII pneumonia.A seasonal variation was observed in the prevalence of pneumococci in the nasopharynx of both contact and control groups ; the highest incidence was noted in the early spring and the lowest during the late summer. This seasonal variation was not a peculiarity of any one type of pneumococcus.There is no evidence from these data that poor eco¬ nomic conditions, with resultant overcrowding, influ¬ enced the distribution of pneumococci. Furthermore, the various specific strains were widely distributed through the various age groups.Our evidence indicates that contacts having colds at the time of contact with a case of pneumonia did not harbor pneumococci more frequently than contacts with no colds. In some instances, however, an epidemic of acute colds in the family did seem to be a factor in the high prevalence of types I and II in contacts.One question in relation to the epidemiology of type I and type II has not been answered by these studies ; namely : Does a patient with lobar pneumonia due to type I or type II actually infect his family contacts, or do these strains invade several members of a family -possibly following or coincident with a family epi¬ demic of colds-with subsequent development of lobar pneumonia in one member of that family? Either alternative might explain the relatively high prevalence of type I and type II pneumococci in family contacts of a case of pneumonia due to these types.In an attempt to determine this point, we are con¬ ducting an intensive study of the families in which cases of type I and type II pneumonia occur. CONCLUSIONS 1. In a study of over a thousand persons, type I and type II pneumococci were found to be much more prevalent in the nasopharynx in immediate family con¬ tacts of...