L. Forró szerk.: Miscellanea Zoologica Hungarica 7. 1992 (Budapest, 1992)
Topál, Gy.; Csorba, G.: The subspecific division of Rhinolophus luctus Temminck, 1835, and the taxonomic status of R. beddomei Andersen, 1905 (Mammalia, Chiroptera)
A few scatterplots of indices were also made. Among these the index of measurements: MANDIB-LE/LC-P4-LE vs C-CONDYL/MAST-WID put the available specimens of R. beddomei close to a specimen of R. I. foetidus (BNHM 59.183.) and but to a single R. I. pemiger (BNHM 91.10.7.55.) because of their clearly shorter lower C-P4 row and greater mastoid width as compared to the rest of material. The indices: BASIL-LE/NAKNOB-W vs M3-M3-WI/C-C-WIDT showed R. beddomei to have a distinctly narrower nasal knob and C-C width narrower than the other specimens of various subspecies of R. luctus have, except the BNHM 91.10.7.55. and the BNHM 21.1.6.2. R. I. pemiger. According to other plots of indices, data of R. beddomei well diverge from those of almost all the others because of their P 2 being narrower (with some overlaps with that of BNHM 21.1.6.3., BNHM 78.2310. and BNHM 91.10.7.55.), then the M r M 3 row being comparatively longer (the closely placed BNHM 7.1.1.294. and BNHM 79.11.21.141. specimens of R. I. pemiger have long C-P4 row). Besides the above specimens, in mastoid width the BNHM 21.1.6.2. and BNHS 3073, in C-C width BNHM 91.10.7.55. and BNHM 21.1.6.2. specimens ofR. I. pemiger, as well as in P 2 width the BNHM 76.9.20.12., in M r M 3 length and in C-C width the BNHM 98.11.3.9. and in C-P 4 length the BNHM 94.9.29.4. and BNHM 76.9.20.12. specimens of R. I. foetidus are near R. beddomei. The individuals of the present study material were clustered by several linkage algorithms using Eucledian distance between them. Many of the dendrograms separate R. beddomei well from the rest of specimens and some of them, to lesser extent, also R. I. morio and R. I. foetidus from R. I. pemiger. For the single linkage clustering method all the thirty specimens were analysed for the following 13 characters: C-CONDYL, TOTAL-LE, BASIL-LE, ZYG-WIDT, MAST-WID, C-C-WIDT, M3-M3-WI, UC-M3-LE, PALBRI-L, MAND-LEN, LC-M3-LE, INTERO-W, and NAKNOB-H. At a distance of 0.686 the three South-Indian R. beddomei were clustered out of all the rest, moreover, at 0.550 the only Malayan specimen was separated. At 0.333 all Bornean and central Thai specimens were put together. To the present purpose the average linkage method proved to be the most suitable one (Fig. 18). All characters of all the 30 specimens were used. At a distance of 1.570 the three R. beddomei are separated off, while the next branching comes at 0.730 for two groups. In one of these a single Burmese, the Malayan, and but two North-Indian specimens are grouped along with all the Bornean R. I. foetidus and the two R. I. morio from Central Thailand. The group of these two subspecies is separated at 0.346 from the northern specimens. When, instead of the total number of variables only 10 were used (TÔTALLE, MAST-WID, C-C-WIDT, NAKNOB-H, UC-M3-LE, PALBRI-L, INTERO-W, M3-M3-WI, MAND-LEN and ZYG-WIDT), the South-Indian species was put together with the Malayan R. I. morio and separated from the others at a distance of 1.686. Within the other group of clusters the subspecies R. I. foetidus and R. I. morio appeared in one cluster at 0.489. In many of the diagrams the Nepalese examples were put side by side and when available, close to the specimen from Sikkim. Interestingly enough, the two animals from N. Thailand were clustering amongst the North-Indian and Burmese R. I. pemiger and certainly far from R. I. morio in Central Thailand.
