Table 1 shows activation indices in 110 cross combinations among 11 species of which 83 were actually examined to give viable seeds (successful cross, marked with +) or inviable seeds (unsuccessful cross, marked with ^*). In the crossing experiments AI indicated a wide variation of 15.9-183.3% as compared with 50% in selfed plants. The results were more briefly shown in a diagramatic illustration (Fig. 1). That is, AI was divided in classes differing by 10%, e.g. the class 50-60% containing AI with 51 to 60%, and so on. There was given the frequency distribution of crosses where abortive one was marked with^* .The AI series was arbitrarily divided into four groups, 10-20, 20-70, 70-90 and 90-190% in which the seed germination was generally variable, stable, variable and nothing, respectively. In the last group only one cross, Ae. squarrosa x ventricosa, was successful though some authors failed to get viable seeds. This is merely to make a special case of the general feature of cross incompatibility. Only a few data were available in abortive crosses of the first and second group. Therefore it has need to check them in detail.

Without a knowledge of detailed seed development it is speculative to make a further discussion in connection with activation index. However, based on a comparative investigation with that of Avena and others, the following is probably assumed as shown in Fig, 1. In general the AI 20-70% give nearly normal or normal seeds (N-type), and more than 90% Produce empty, shrivelled seeds (E-type). The AI 70-90% show an intermediate type of seeds (Ps-type) between them. In the AI 10-20% small and weakly developed seeds are obtained and their germination is usually low (W-type).

The polar-nuclei activation hypothesis is almost successfully applied to the seed abortion in interspecific hybridizations of Aegilops. The result of some unexamined crosses in Table 1 will be expected by their activation indices in advance.

The supplement of some appropriate crossing data, indicating hybrid-seed development and germination (%) would assist the establishing of a more perfect cross-incompatibility system in the genus Aegilops.

Literature Cited

BERG, K.H.v. 1937. Beitrag zur Genomanalyse in der Getreidegruppe, Zuchter 9 : 157-163.

KATAYAMA, Y. 1933. Crossing experiments in certain cereals with special reference to different compatibility between the reciprocal crosses. Memoirs Coll. Agr., Kyoto Imp. Univ. 27 : 1-75.

KIHARA, H. 1937. Genomanalyse bei Triticum und Aegilops, VII. Kurze Ubersicht uber die Ergebnisse der Jahre 1934-1936. Memoirs Coll. Agr., Kyoto Imp. Univ. 41 : 1-61.

KIHARA, H. 1949. Genomanalyse bei Triticum und Aegilops, IX. Systematischer Aufbau der Gatung Aegilops auf genomanalytischer Grundlage. Cytologia 14 : 135-144.

KIHARA, H. und F. LlLIENFELD 1932. Genomanalyse bei Triticum und Aegilops, IV. Untersuchungen an Aegilops x Triticum- und Aegilops x Aegilops-Bastarden. Cytologia 3 : 384-456.

KIHARA, H. and M. TANAKA 1954. Genetic stock at Laboratory of Genetics. Kyoto University, Japan.