made in groups with differing severities of illness or temperatures intergroup comparisons of the means of the logarithmic values were made by Student's unpaired t test. Case 4 was excluded from statistical analyses involving gastrin concentrations.

No abstract

Experience with a computer in an obstetric department is presented. A regular statistical analysis of the work of the department was readily available and computer printed summaries were produced for 86 per cent of the patients within one week of their discharge home. Solutions to problems encountered are discussed. It is concluded that a computer can be used economically and advantageously in this way even in hospitals without a computer on site. Future developments in the use of the computer are outlined. AND

No abstract

No abstract