1. Eighty-five samples from fifteen different legume seed lines generally available inthe UK were examined by measurements of their net protein utilization by rats and by haemagglutination tests with erythrocytes from a number of different animal species. From these results the seeds were classified into four broad groups.
2. Group a seeds from most varieties of kidney (Phaseolus vulgaris), runner (Phaseolus coccineus) and tepary (Phaseolus acutifolius) beans showed high reactivity with all cell types and were also highly toxic.
3. Group b, which contained seeds from lima or butter beans (Phaseolus lunatus) and winged bean (Psophocarpus tetragonolobus), agglutinated only human and pronase-treated rat erythrocytes. These seeds did not support proper growth of the ratsalthough the animals survived the 10 d experimental period.
4. Group c consisted of seeds from lentils (Lens culinaris), peas (Pisum sativum), chick-peas (Cicer arietinum), blackeyed peas (Vigna sinensis), pigeon peas (Cajanus cajan), mung beans (Phaseolus aureus), field orbroad beans (Vicia faba) and aduki beans (Phaseolus angularis). These generally had low reactivity with all cells and were non-toxic.
5. Group d, represented by soya (Glycine max) and pinto (Phaseolus vulgaris) beans, generally had low reactivity with all cells but caused growth depression atcertain dietary concentrations. This growth depression was probably mainly due to antinutritional factors other than lectins.
6. Lectins from group a seeds showed many structural and immunological similarities. However the subunit composition of the lectin from the tepary bean samples was different from that of the other bean lectins in this or any other groups.