The proximate composition of garlic (cv. Early Californian) was determ
ined and compared to values reported in the literature. The specific h
eat and thermal conductivity were determined using the method of mixtu
res and a modified Pitch apparatus, respectively and both approaches p
roved sufficiently accurate for such purposes. An analysis of variance
revealed that moisture content had a highly significant effect on bot
h specific heat and thermal conductivity of garlic. A linear empirical
model was developed to characterise the specific heat as a function o
f moisture content. The model predictions were very close to values pr
edicted on the basis of the proximate analysis. The measured thermal c
onductivity data were fitted to parallel, perpendicular and random str
uctural models as well as empirical mathematical expressions. The para
llel and exponential models gave satisfactory agreement with the data.