REVIEW OF ALGERIAN MACROPROTODON 



87 



the deposition of these specimens is not indicated. M. c. brevis, with 

 which the western population of M. c. mauritanicus (sensu Wade. 

 1988) is synonymised by Busack & McCoy (1990), is similarly 

 poorly represented in collections from areas where it would meet 

 neighbouring taxa. A full revision of M. c. cucullatus and M. c. 

 brevis is postponed pending study of material from eastern and 

 southern Morocco. 



The subspecies are accorded full species rank. The balance of 

 evidence suggests that the distinctness between them is maintained 

 with little suggestion of any intergradation between the taxa. 



MATERIALS AND METHODS 



The material examined in the course of this work is listed at the end 

 of each species treatment. Abbreviations used are as follows: Acad- 

 emy of Natural Sciences, Philadelphia (ANSP); Natural History 

 Museum, London (BMNH); California Academy of Sciences, San 

 Francisco (CAS); Carnegie Museum of Natural History, Pittsburgh 

 (CM); Field Museum of Natural History, Chicago (FMNH); Mu- 

 seum of Comparative Zoology, Harvard (MCZ); Museo Nacional de 

 Ciencias Naturales, Madrid (MNCN); Museum d'Histoire naturelle. 

 Geneve (MHNG); Museum National d'Histoire Naturelle, Paris 

 (MNHN); Naturhistorisches Museum Basel (NMB); Rijksmuseum 

 van Natuurlijke Historic Leiden (RMNH); Senckenberg Museum. 

 Frankfurt am Main (SMF); University of Colorado Museum, Boul- 

 der (UCM); National Museum of Natural History. Washington, D.C. 

 (USNM); Museum fiirNaturkun.de, Universitat Humboldt. Berlin 

 (ZMB); Zoologische Staatssammlung, Mlinchen (ZSM); 

 Zoologisches Museum der Universitat Hamburg (ZMH); 

 Naturhistorisches Museum, Wien (NMW); personal collection of 

 the author (EW). 



The primary source for Fig. 7 has been the bioclimatic map of 

 Emberger et al. (1962) from which relevant data has been extrapo- 

 lated and forms the substance of Fig. 1 2. Other maps have also been 

 consulted. The bioclimatic map utilises meteorological data and 

 quantifies dry season drought taking into account not only tempera- 

 ture and precipitation but also atmospheric humidity as the 

 xerothermic index (x) or number of physiologically dry days in the 

 dry season (Emberger et al., 1962, pp. 12-19). The area under 

 consideration broadly translates into north-south divisions of seven 

 subregions with their respective xerothermic indices ranging from, 

 in the more northern latitudes, the sub-mediterranean (.v =<40) to 

 those in the south, the desertic [x =300+). The xerothermo- 

 mediterranean subregion (x = 150-200) is seen as a transitional zone 

 between the Mediterranean and the Arid-Saharan regions. Translit- 

 eration of indigenous place names mainly follows the Michelin 

 (tourist) map 958 'Algerie-Tunisie'. The convenient symbols of 

 Busack and McCoy (1990) have been used. 



Each synonymy includes most of the generic and specific combi- 

 nations. Pages refer to citations of the taxa. Most of the references in 

 which either sources are not given or the species is just mentioned, 

 have been excluded. 



CHARACTERS EXAMINED 



All specimens were sexed. Morphological characters used by Wade 

 (1988) and those used and considered by Busack & McCoy (1990) 

 were re-evaluated. Of those of the latter the sum of the ventral + 

 subcaudals, anterior temporals and postocular counts were not 

 analysed; all their other characters were found to be useful in 



varying degrees. Additional characters such as maxillary tooth 

 counts and scale row reductions were incorporated into the analysis. 



Dorsal scales rows. These are normally understood to mean those 

 counted arounud midbody. The reductions were recorded in the 

 manner proposed by Dowling ( 195 lb). They occur behind the head 

 from 25 in four stages to 19 rows until anterior to the vent where the 

 count falls to 17 or more rarely to 15. Occasionally after 19 there 

 may be no further reduction. These are mostly due to fusion between 

 the 4th and 5th rows but those of the 4th and even the 2nd and 3rd are 

 not infrequently involved. Exceptionally, especially in individuals 

 with aberrant dorsal scalation. fusion may occur in other rows, in 

 particular the vertebral and paravertebral. For practical purposes 

 only the reduction from 21 to 19 on the neck and 19 to 17 anterior to 

 the vent, not taking into account any irregular fluctuations, have 

 been considered. The length of the head is defined in this work as the 

 distance from the snout to the retroarticular process of the mandible. 

 The 1st ventral is situated at or slightly posterior to this point in 

 vertical alignment. 



In addition to dorsal counts at midbody Busack & McCoy (1990) 

 considered counts at a point one head length (not specified) behind 

 the head and one of same value anterior to the vent. These characters, 

 especially the latter, were found to be significant. The histograms 

 (Figs. 8 & 9) show (ordinate) each point of reduction at the relevant 

 ventral scute expressed as a percentage of the total ventral number. 

 A mean to the nearest integer was taken when the reductions were 

 unequal on either side. Subsequent fusions and divisions after the 

 main points of reduction were not quantified. 



Ventral and subcaudal counts. These were counted in the manner 

 of Dowling (1951a) and are summarised in Table 1. Addition of the 

 ventrals to the subcaudals (Busack & McCoy, 1990) whilst not 

 completely eliminating sexual dimorphism did. nevertheless, 

 diminish it, whereas as discrete states they revealed clear sexual 

 dimorphism in all the taxa (Fig. 11). With the exception of M. c. 

 cucullatus (sensu Wade, 1988) interpopulational variation in these 

 characters was usually slight or too discordant to permit determination 

 of any geographic trends in either of the northern forms. 



Supralabials. Nearly all the material examined here possessed 8 

 supralabials on each side of which the 4th & 5th meet the eye: one 

 individual of M. mauritanicus had 9 on both sides. 



Supralabial-parietal contact. (Table 2). In the North African 

 material this character was useful in differentiating M. mauritanicus 

 from M. cucullatus . 



Infralabials. (Table 2). The anterior series, i.e. those in contact 

 with the genials, normally numbered 6, there being few departures. 

 Significant differences were found in the posterior or 'free' series of 

 infralabials. 



Maxillary teeth. (Table 2). The maxilla bears a series of six teeth 

 increasing in size followed by a diastema and a second numbering 

 from three to five of smaller size that precede the fangs. Dumeril and 

 Bibron ( 1 854) gave the number of maxillary teeth as '6+3 ou 5+2' 

 which is understood here to mean 6, followed by 3 or 5+II. Significant 

 differences occurred in the second series of teeth, i.e. that which 

 precedes the fangs. 



Head pattern. The variability in the elements of head pattern first 

 noted by Boulenger has been analysed by Wade (1988) and Busack 

 and McCoy (1990). From the confusing array of configurations in 



