Letter: Skin reactions to practolol.

They compared 15 "cot death" cases with 15 "controls." The ages of the cot death patients ranged from 11 days to 2 years, though most workers in this field agree that the sudden infant death syndrome probably does not occur before 4 weeks of age and is rare after 9 months.' Seven of the 15 control cases died within four days of birth, and since they may have been subject to postnatal respiratory problems they cannot be considered comparable with a series of cot deaths. It was stressed that no evidence of degeneration was found in the atrioventricular bundle. However, remoulding of the junction between the atrioventricular node and the bundle of His where it penetrates the central fibrous body, an area we term the isthmus, does occur.2 3 Furthermore, such remoulding is not present at birth but is apparent within 2-3 weeks.4 Perhaps this reshaping of the atrioventricular conducting tissue is not apparent by the method of sectioning the heart described by Dr. Anderson and his colleagues, but it is easily seen if the complete heart is processed intact and serially sectioned by the technique of Ferris and Maclennan.5 The atrioventricular node receives its blood supply from the right coronary artery or the circumiflex branch of the left and the His bundle from the anterior descending branch of the left coronary artery. It is interesting to note that this remoulding occurs at the junction of these separate blood supplies. Haemorrhages were reported in both the cot death and control series. However, no cause of death was given in the control series, and since we contend that petechial haemorrhages in the heart occur in death associated with terminal hypoxia and a rapid change in intrathoracic tension it is important that any terminal hypoxic episode should be recorded. There is much evidence6 in favour of an acute respiratory death associated with virus infection in some oot deaths. In a partly completed study of 38 cot deaths and 30 control cases we have found petechial haemorrhages in the hearts of 18 of the cot death cases and in only two of the controls (see table). One of the two control cases with cardiac haemorrhage died from massive inhalation of vomit and the other from acute myocarditis. In a series reported by Ferris7 haemorrhages were found within the atrial wall, and while the histological 'identification of internodal tracts is disputed there is considerable physiological evidence for the existence of such tracts.8 Haemorrhages close to these tracts may be an important factor in the failure to survive in cases of what might otherwise be acute recoverable hypoxia. In one of the cases illustrated by Dr. Anderson and others (their fig. 2) there was a history of cardiac arrest followed by cardiac massage. The illustration of the atrioventricular node shows aibnormal vascularity and an unusual pattern of nodal tissue similar to that seen in hypertrophic