Diagram illustrating a principal co-ordinates analysis (Gower, 1966; 1967) of the definitive tobamoviruses computed from the amino acid compositions of their coat proteins. Analyses for 75 isolates were classified (MULTCLAS program using Euclidean metric; Lance & Williams, 1967a, b) and fell into 10 clusters or individuals. The mean compositions of these clusters or individuals were then used to compute similarity measures whose principal co-ordinates were calculated. The cluster called TMV contained 19 isolates (type, 01, 03, 04, 05, 06, 0M, M, J14D1, YA, GA, Mosch 6 & 19, Mutant 470 & 483, PM1D, PM21B, PM4A, PM4C), the CGMMV cluster 2 isolates (CGMMV, CV3 Jap), the CV4 cluster 7 isolates (4 of CV4, CV3 Berk, CV4 Czech, CV4 Berk), the ORSV cluster 3 isolates (ORSV, 02, 07), the HRV cluster 2 isolates (HRV, Lychnis), the T2MV cluster 7 isolates (U2, G-TAMV, HNO2 262 & 328, P0249, NBSI 206 & 223), and the ToMV cluster 32 isolates (Y-TAMV, dahlemense, NBSI331, Mosch 1-5, 7-10 & 12-18, VC6O, VC61, AC9, HD, SJ, VEN, Aus-11, Dut-1, YLGP, K-1, PTA, PTV, SAF). The sources of data were Chessin et al., 1967; Desc. 153; Francki et al., 1971; Hariharasubramanian et al., 1973; Kado et al., 1968; Kurachi et al., 1972; Linnasalmi & Toiviainen, 1974; S. J. Morris, pers. comm.; Mosch et al., 1973; Nozu et al., 1971; Rombauts & Fraenkel-Conrat, 1968; Tsugita, 1962; Tung & Knight, 1972; van Regenmortel, 1967; and Wang & Knight, 1967.
The relative positions of the ‘individuals’ in the third dimension of the ordination are indicated by their sizes. The first dimension contains 32% of the ‘comparative information’ of the ordination, the second contains 21% and the third 20%; altogether 73%.