P.D. JENKINS, C.J. RAXWORTHY AND R.A. NUSSBAUM 



reduced inM. soricoides. In the mandibular dentition ofM. soricoides, 

 both i3 and p2 are very reduced and the latter is unusual in having a 

 single root; both of these teeth in M. fotsifotsy are reduced but less 

 markedly so and p2 has two roots. The lower canine is similar in both 

 species but p3 is more caniniform in M. soricoides. The third lower 

 molar shows a similar degree of reduction of the talonid in both 

 species. 



The dentition of the new species is distinctive and does not readily 

 group with any of the species clusters described by MacPhee (1987) 

 but, as outlined above, does show some similarity to M. soricoides, 

 which may represent a separate cluster. The degree of development 

 of accessory cusps on II— C is similar to that of members of the 

 cowani cluster but less marked than in the longicaudata cluster, also 

 the presence of interproximal gaps between the upper incisors 

 occurs in M. fotsifotsy and the cowani cluster. In contrast, M. 

 fotsifotsy is more similar to the longicaudata cluster in that the 

 buccal aspects of P3 and P4 are similar in shape, the lower canine 

 lacks a paraconid and the crown of p2 appears anteroflexed as in c, 

 due to the shorter convex anterior slope and longer concave posterior 

 slope. The relative proportions of the anterior teeth of M. fotsifotsy 

 differ from either the cowani or the longicaudata cluster, they are 

 similar to, but less extreme than those in M. soricoides, as shown in 

 Table 3. 



Table 3 Relative proportions of the anterior teeth of Microgale fotsifotsy 

 in comparison with M. cowani cluster, M. longicaudata and M. 

 soricoides. 



cowani cluster 

 longicaudata 

 fotsifotsy 

 soricoides 



I1>I2>I3<C>P2< 

 I1>I2>I3<C>P2> 

 II >I2 » 13 <C »P2 < or; 

 II » 12 >» 13 <« C>»P2<« 



cowani cluster il=i2>i3<c = 



longicaudata il < i2 > i3 < c > 



fotsifotsy il > i2 » i3 < c » 



soricoides il » i2 » i3 « c »> 



or 



P3 



CsP3 



P3 



C>P3 



; P3 



C>P3 



P3 



C>P3 



< p2 < p3 



c <p3 



p2< p3 



c = p3 



p2<p3 



c>p3 



p2<« p3 



c sp3 



Microgale drouhardi G. Grandidier, 1934 



Microgale melanorrhachis Morrison-Scott, 1948 

 Microgale cowani Thomas: MacPhee, 1987, in part. 



Holotype. MCZ 45034 (specimen A in original description) 

 juvenile female, body in alcohol, skull extracted, collected by 

 Monsieur Drouhard. 



Type LOCALITY, environs of Diego-Suarez[Antsiranana,c. 12°16'S 

 49° 18'E-see MacPhee, 1987]. 



Paratypes. MCZ 46007-4601 1 (specimens B-F in original de- 

 scription), MCZ 460 12 (juvenile mentioned in original description), 

 all in alcohol. All from the same locality as the holotype. 



Referred material, environs of Diego-Suarez [Antsiranana, c. 

 12°16'S 49°18'E]: MCZ 46013^6019. 



PN de la Montagne d' Ambre, 5.5 km SW of Joffreville [Ambohitra], 

 12 o 31'S49 o 10'E, 1000m, in relatively undisturbed rain forest: FMNH 

 154489; FMNH 154491; FMNH 154493-154494; FMNH 154499; 

 FMNH 154514-154515; FMNH 154517; FMNH 154561-154564; 

 FMNH 154567-154571. 



Camp 1, Antomboka River, PN de la Montagne d' Ambre, 

 Antsiranana Fivondronana, Antsiranana Province, 1 2°32'S 49° 1 0'E, 

 1150-1250m, rain forest: UMMZ 171000-171015; BMNH 

 1996.280; BMNH 1996.281. 



5km S of Joffreville [Ambohitra], Mont[agne] d' Ambre, Diego 



Suarez [Antsiranana] Province, [c 12°32'S 49°10E], rain forest: 

 USNM 341692; 6km S of Joffreville [Ambohitra], Mont[agne] 

 d' Ambre, Diego Suarez [Antsiranana] Province, [c 1 2°32'S 49° 1 0'E] , 

 rain forest: USNM 341693. 



Bekolosy, RS de Manongarivo, Ambanja Fivondronana, 14°03'S 

 48°18'E, 1150m, rain forest: UMMZ 171016-171019. 



Matsabory, RNIdeTsaratanana, Ambanja Fivondronana, 14°09'S 

 48°58'E, 2350m, rain forest: UMMZ 171020. 



RS d' Ambatovaky, Soanierana-Ivongo Fivondronana, Toamasina 

 Province 16°51'S 49 o 08'E-49°16'E, 360-600m: BMNH 91.220; 

 BMNH 91.221-91.226. 



RNI de Zahamena, Ambatondrazaka Fivondronana 17°41'S- 

 17°42'S 48°46'E, 850-920m: UMMZ 171021-171027. 



RNI de Zahamena, Vavatenina Fivondronana 17°44'S 48°59'E, 

 420-560m: UMMZ 171028-171039. 



PN de Mantady, Moramanga Fivondronana, Toamasina Province, 

 18°51'S 48°27'E, 1100m, rain forest: UMMZ 168471-168474. 



Didy, E of Lake Alaotra [Toamasina Province, c 1 8°02'S 48°32'E 

 - see Carleton & Schmidt, 1990]: USNM 328686. 



Perinet [= Andasibe], near Moramanga, eastern Madagascar, 

 19°00'S 48°30'E, 3000 feet [915m]: BMNH 48.88 (holotype of 

 Microgale melanorrhachis). 



43 km S of Ambalavao, junction of Sahanivoraky and Sahavatoy 

 Rivers, RNI d' Andringitra, 22° 1 3'S 47°00'E, 8 1 0m: FMNH 1 5 1 627; 

 FMNH 151756. 45 km S of Ambalavao, E bank of Iantara River, 

 along Ambalamanenjana-Ambatamboay Trail, edge of RNI 

 d' Andringitra, 22° 1 3'S 47°01'E, 720m: FMNH 151626. 



6 miles E of Ivohibe, 22°30' 47°00'E, 5000 feet [1525m], [high 

 rain forest]: BMNH 48.87. 



TAXONOMIC NOTES. The original description of M. drouhardi was 

 based on several specimens, consisting of whole bodies preserved in 

 alcohol, with the exception of the holotype, which is a body with the 

 skull extracted. Grandidier (1934) believed that all but one of these 

 specimens were adult but MacPhee (1987) demonstrated that the 

 holotype is immature with the dentition partially deciduous (Stage 2, 

 see MacPhee, 1987: 13). As MacPhee emphasised, repeated 

 misidentifications and faulty taxonomic judgements have occurred 

 in this genus because of the problems of distinguishing between 

 deciduous and permanent dentitions. However, two of the speci- 

 mens mentioned in the type description are adult and, more critically, 

 another specimen (MCZ 46017) prepared as a skull and skeleton, is 

 adult and, although not mentioned in the original description, was 

 evidently available to Grandidier. Grandidier deposited his private 

 collection, including the original series of M. drouhardi, in the 

 Museum of Comparative Zoology, Harvard. In effect, this material 

 became unavailable to subsequent authors, such as Morrison-Scott 

 (1948) and Heim de Balsac (1972), who were unaware at which 

 institution in the the United States these specimens were held. 

 MacPhee was apparently the first author to re-examine this material, 

 he concluded that, although at the extreme of the size range, the 

 morphology of the teeth was such that the specimens were insepara- 

 ble from M. cowani and that M. drouhardi should be treated as a 

 synonym. 



The type locality of 'the environs of Diego-Suarez [Antsiranana]' 

 is unfortunately vague and, as pointed out by MacPhee (1987) and 

 Nicoll & Rathbun (1990), includes ecologically diverse habitats of 

 dry forest on limestone at Tendrombohitr' Antsingy, Ankarana, 

 Analamera and Cap d' Ambre, grassland and dry forest on the lower 

 slopes of the basaltic Montagne d' Ambre, with rain forest at higher 

 altitudes of Montagne d'Ambre. Since most species of Microgale 

 are recorded from rain forest (M. pusilla Major, 1896b and M. 

 principula Thomas, 1918 contained in owl pellets of indeterminate 



