The hypothesis that the data of a sample came from a normal population can be tested by a goodness of fit procedure. Often, however, the departures from normality are evaluated visually. For this, you can plot the data on a scale of normal probability (using a sheet of paper with this scale printed on it or using a computer program). The subject is discussed in books on biostatistics – e.g. ‘Biometry’ (Sokal & Rohlf) and ‘Biostatistical analysis’ (Zar).