From tubers of 34 varieties of Solanum tuberosum, extracts were prepared and analysed for crude protein and coagulable protein by the Kjeldahl method. Content of coagulable protein was 0.37 1.24 '7,,. From the same material, juice was prepared with a juice centrifuge and analysed for coagulable protein by the Kjeldahl and microbiuret method. The average amount of coagulable protein in the juice was 81.8 ~ of the total coagulable protein. The correlation coefficient between coagulable protein in the juice and total coagulable protein was 0.956**. The correlation coefficient between Kjeldahl and microbiuret data for coagulable protein in the juice was 0.956**. Analysis of tuber juice by the microbiuret method is recommended as a rapid screening technique for coagulable protein.
Relationships between protein data of the 34 varieties and earliness, yield and content of dry matter were analysed statistically. Content of coagulable protein in fresh material correlated with content of dry matter (r 0.756**), yield of fresh potatoes (r 4).615**) and earliness (r 0.361") but not significantly with yield of dry matter (r 0.309). Coagulable protein in dry matter correlated with fresh yield (r -0.525**), but not significantly with content of dry matter (r 0.260), yield of dry matter (r 0.131 ) and earliness (r 0.054). Path coefficient analysis showed that 67 '~ of the variation in coagulable protein in fresh material was statistically determined by earliness, content of dry matter, fresh yield and content of uncoagulable protein in fresh material, whereas only 34~o of the variation in content of coagulable protein in dry matter was determined by these components. A high content of coagulable protein in fresh or dry matter can be combined with early maturity and with high yield of dry matter. Coagulable protein in dry matter seems to be a more suitable criterion of selection than coagulable protein in fresh material * P ~< 0.05: ** P ~< 0.01.