Vörös A. szerk.: Fragmenta Mineralogica Et Palaentologica 11. 1983. (Budapest, 1983)

above M

2

, C-M

3 length (crowns), C-M

3 length (alveoli), C-P

4 length (alveoli), C-P

4 length (crowns), P

4-M

3 length (crowns), M^-M

3 length (crowns), C alveolus length, C alveolus width, combined p2_p3 alveolar length, cross-section length of P

4

, cross-section width of P

4

, anteroposterior length of M*, width of between lingualmost and metastylar margins, width of Ml between lingualmost and parastylar margins, antero-posterior length of M

2

, width of M

2 between lingualmost and metastylar margins, width of M

2 between lingualmost and parastylar margins, antero-posterior length of M

3

, width of M

3 between lingual and parastylar margins. Besides the above measurements used for computation, the following ones also have been taken and included in this paper. Total length of skull, condylobasal length, zygomatic width, braincase width, braincase length, palatal width beyond toothrows, C crosssection length, C cross-section width, P

2 cross-section length, P

2 cross-section width, P

3 cross-section length, P

2 cross-section width, length of mandible, height of mandibular ramus under Mj, height of coronoid process from the basal sinus, C-M3 length, C-P4 length, P4-M3 length, M1-M3 length, cross-section length of lower C, cross-section width of lower C, cross-section length of P3, cross-section width of P3, cross-section length of P4, cross-section width of P4, antero-posterior length of Mj, M\ talonid width, anteroposterior length of Mg, Mg talonid width, antero-posterior length of M

3

, M3 trigonid width, M 3 talonid width. VAN VALEN'S (1966) anatomical nomenclature was used for the dental particulars. Principal component solution of factor analysis was carried out using programme BMDP4M (DIXON & BROWN, 1981). RESULTS AND CONCLUSIONS In the principal components analysis of the 37-character correlation matrix, a total of 46.1 % of the sample variation was accounted for by the first two principal components. These two components describe major differences among the groups (Fig. 1). The first principal component separated the M, exilis - janossy sp. n. complex and M. helleri

. As can be seen in the figure, members of this complex cannot be separated based on the characters used in the analysis. The second principal component separated specimens of exilis - janossy sp. n. and helleri from paradaubentoni sp. n. Major character loadings on the first principal component were P^M^ length (0. 857), M

1-M

3 length (0. 854), M

3 width (0.852), width of M

2 between lingualmost and metastylar margins (0.850), rostral width at M

3-M

3 (alveoli) (0.846), rostral width at M

3-M

3 (crowns) (0.842), C-M

3 length (crowns) (0.819), width of M between lingualmost and parastylar margins (0.813), width of M* between lingualmost and parastylar margins (0.804). High loadings on the second principal component were for C alveolus width (0. 639), strength of M

2 paraconule (0. 532), strength of M* paraconule (0.515), strength of M

3 paraconule (0.478), rostral width at C-C (alveoli) (0. 500), width of narial notch (0. 482), length of anterior palatal emargination (0.459), combined alveolar length of P

2

P

3 (-0.621), C-P

4 length (crowns) (-0.582). The mandibles of M. estramosensis sp. n. and of M. cf. blythi were treated separately. The new species is comparable only with M. mystacinus and the Asiatic siligorensis and caliginosus

. The M. cf. blythi is a closely related fore-runner of M. blythi

. Rostrum and upper dentition of M. helleri Kowalski HELLER (1936) in his original description did not mention any details of the skull or the upper dentition of this species. At Osztramos Localities 9 and 13 this species comprises the so called small-sized species group of Myotis

. As a previous computer analysis has shown (TOPAL in press), it is rather separated from all other fossil bats found at both Osztramos and Podlesice, and from all recent animals, according to the Czekanowski's similarity measure. Though, because of its smaller size, it seemed to be related to other small Myotis

. Actually, it could be compared to M. danutae Kowalski, 1956 from Podlesice, and the latter itself is somehow related to _M. emarginatus (GEOFFROY, 1806). Its full skull is not known, but there is a rather well preserved rostrum among other cranial fragments and mandibles. The rostrum is a very gracile, elongated struc-