taxonID	type	description	language	source
03EB87D5FF86FF83FDB5CA08FB46AA29.taxon	diagnosis	Diagnosis (adapted from Hoffman, 1982; Golovatch, 1996, 2001). Small-bodied Polydesmidea incapable of enrolling into a spiral, usually colourless or black, 3 – 15 mm in length with 19 or 20 segments, including epiproct, as adults; antennae with article 5 longest; collum enlarged and flabellate, usually partly or completely covering head in dorsal view, anterior margin indented to varying degrees forming scallops or lobes; paranota large, margins incised or lobed, at least some ozopores located on elongated, tubular ‘ porosteles’; metaterga typically roughened and without, or with only inconspicuous, setation, surface pustulate or tuberculate to varying degrees, paramedian tubercles often enlarged into subparallel longitudinal crests. Distribution in the USA. Northern, central and southern Texas; western and northern Louisiana through north-central Alabama to coastal Georgia and southward throughout peninsular Florida and the Keys; greenhouses in Illinois (figures 1, 2); also occurring in Hawaii, Puerto Rico and the US Virgin Islands. Golovatch (1996) mapped the family’s distribution worldwide with Florida, the Gulf Coastal fringe and extreme southern Texas shaded. This area is generally accurate, but it extends farther northward in Alabama, Louisiana and Texas, and pyrgodesmids have not been encountered in southeastern Texas. I regard the absence of records from the western Florida panhandle as an artifact reflecting lack of collecting and shade the entire state in figure 1. Remarks. Hoffman (1976) noted the synonymy of Stylodesmidae with Pyrgodesmidae and that the latter holds priority. The widespread occurrence of parthenogenesis in the Pyrgodesmidae has been noted by Shelley and Golovatch (2001), Golovatch and Sierwald (2001) and Golovatch et al. (2001), and it should be reiterated that this is true of US populations of Poratia digitata, P. obliterata and Calyptodesmus sanctus. No males have ever been collected here of these species. It is possible to confuse the ornamented, pustulate, small-bodied polydesmidan, Prosopodesmus jacobsoni Silvestri (Haplodesmidae), introduced into Florida and Louisiana (Shelley and Golovatch, 2000), with representatives of the Pyrgodesmidae. The following key distinguishes these taxa and representatives of the Pyrgodesmidae in the continental USA. Key to the Pyrgodesmidae in the Continental USA 1 Paranota at most only slightly declined, extending generally horizontally, subparallel to substratum; ozopores positioned at paranotal margins; Georgia and Florida to Texas and the Rio Grande, Illinois (native and introduced) ... Pyrgodesmidae: 2 – Paranota strongly declined, extending nearly perpendicularly, directed ventrad (downward) toward substratum; ozopores removed from paranotal margins; Florida, Louisiana (introduced) ..... Haplodesmidae: Prosopodesmus jacobsoni Silvestri 2 Scallops on collum completely and widely separated from each other, except for two lateral-most, which are fused basally and slightly indented (figure 3) ...... ................. Poratia Cook and Cook: 3 – Scallops on collum fused basally to varying degrees (figures 4 – 8) ...... 4 3 Adults with 19 segments, largely with three lobes on paranota of segments 16 – 18; Florida, Louisiana, Texas, Illinois ........ Poratia digitata (Porat) – Adults with 20 segments, largely with four lobes on paranota of segments 16 – 18; Georgia .............. Poratia obliterata (Kraus) 4 Porosteles present on segment 9 (potentially present on segments 5, 7, 9, 10, 12, 13, 15 and possibly 16) .................. 5 – Porosteles absent from segment 9 (potentially present on segments 5, 7, 10, 13, 15 and possibly 16) ............. Myrmecodesmus Silvestri: 7 5 Segment 16 with porosteles; poriferous paranota with one broad, subtruncated lobe anterior to porosteles (figure 12); Florida .. ‘ Lophodesmus ’ bituberculatus Loomis – Segment 16 without porosteles; poriferous paranota either with one narrowly rounded lobe anterior to porosteles or two lobes (figures 9 – 11) ......... 6 6 Porosteles located at caudolateral corners of paranota; porostele bearing segments mostly with two lobes anterior to porostele, segment 5 with only one (figure 10); Florida ................ Psochodesmus crescentis Cook – Porosteles located between lobes near mid-lengths of paranota; porostele bearing segments with one lobe anterior to porostele, segment 5 also with one (figure 9); Florida, Georgia, Alabama ........ Calyptodesmus sanctus Schubart 7 Paramedian tubercles coalesced into moderate-size crests, diminishing in height on segments 19 – 20, extending only slightly beyond level of epiproct, latter visible in dorsal view between carinae (figures 17, 18); Texas .. Myrmecodesmus reddelli, new species – Paramedian tubercles coalesced into strong crests, greatly elevated above tergal surfaces on all segments, clearly extending beyond level of epiproct (figures 13 – 16) ... 8 8 Caudalmost tubercles on segment 19 narrowly separated, obscuring epiproct in dorsal view (figure 13); Louisiana, Texas .... Myrmecodesmus formicarius Silvestri – Caudalmost tubercles on segment 19 widely separated, epiproct visible between them in dorsal view (figure 15); Florida to Louisiana .. Myrmecodesmus digitatus (Loomis)	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF83FF8DFD54CDD7FCA5A9D5.taxon	diagnosis	Diagnosis. Nineteen segments including epiproct. Preserved specimens thin, fragile and lightly sclerotized, colour whitish. Scalloped margin of collum upturned to varying degrees, usually strongly; scallops, except for two lateralmost on both sides, completely segregated from each other, usually widely, lateralmost ones fused basally (figure 3). Paranota 2 with three lateral lobes, remaining non-poriferous paranota with three lobes. Porosteles present on segments 5, 7, 9, 10 and 15, positioned at caudolateral corners of paranota, with one anterior lobe. Variation. The degree of upturn of the collum margin varies as does the degree of separation of the adjacent lobes, and individuals with less upturn and separation can be mistaken for Psochodesmus crescentis because of their similar paranotal configurations, although there is only one lobe anterior to the porostele rather than two. In such a tiny milliped, it is sometimes hard to discern the lobes on the collum; I find the most definitive feature of Poratia to be the distinctively different lateralmost lobes, which are fused basally and shallowly segregated. This feature diagnoses Poratia, and the number of segments determines the species. Ecology. Habitat information on vial lables in this study include ‘ in trash pile’, ‘ along railroad’ and ‘ cypress sawdust pile’. Overall distribution. The type locality is a greenhouse in Sweden (Porat, 1889). Poratia digitata also occurs in the US Virgin Islands, Costa Rica, Panama and Indonesia, greenhouses in Paris, France, and Germany, and botanical gardens in Germany, Denmark, Switzerland and Great Britain (Attems, 1899; Silvestri, 1923; Schubart, 1947; Loomis, 1961, 1964, 1968, 1970; Blower, 1986; Pedroli-Christen, 1993; Adis et al., 2000; Golovatch and Sierwald, 2001). Adis et al. (2000), Golovatch and Sierwald (2001) and Shelley and Golovatch (2001) reported that P. digitata is an obligatory greenhouse species in Europe and is presumed to be native to Central America / the Neotropical realm. Occurrence in the USA. Introduced into and established in urban environments in peninsular Florida from Alachua County southward to the Keys; known sporadically elsewhere in association with greenhouses and plant nurseries as far north as northern Illinois (figures 1, 2). Poratia digitata also occurs in the US Virgin Islands (Loomis, 1970; Golovatch and Sierwald, 2001; Shelley and Golovatch, 2001); Golovatch et al. (2001) stated that P. digitata is free-living in the southern USA but occurs strictly in greenhouses in the north. The population in the Bahamas is also free-living. Published records. Florida: Florida in general (Adis et al., 2000). Dade County (Co.), Miami (Loomis, 1973; Shelley and Golovatch, 2001). Illinois: Cook Co., Chicago, greenhouse at Lincoln Observatory (Golovatch and Sierwald, 2001). New records. Bahamas: Andros I., 0.6 mi (1 km) S Stafford Cr., Forfar Field Sta., 3 X, October 1980, NCSM Expedition (NCSM). New Country Record.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF83FF8DFD54CDD7FCA5A9D5.taxon	materials_examined	USA: Florida: Alachua Co., Gainesville, University of Florida Pharmacy Gardens, ca 15 X, 4 November 1959, P. C. Drummond (FSCA). Dade Co., Miami, 5355 SW 92 nd St., ca 10 X, November to December 1972, H. F. Loomis (FSCA, NCSM) and Matheson Hammock, 3 X, 4 July 1977, R. M. Shelley (NCSM); South Miami, Chapman Field, 4 X, 30 September 1928, collector unknown (NMNH). Hillsborough Co., Plant City, 3 X, 2 December 1941, A. B. Gurney (NMNH). Manatee Co., Bradenton, 4 X, 2 December 1966, C. J. Bickner (FSCA). Monroe Co., Lower Matecumbe Key, X, 16 December 1959, R. E. Woodruff (FSCA). Pinellas Co., nr Tarpon Springs, X, 29 September 1959, R. E. Woodruff (FSCA); and Largo, Indian Rocks Nursery, 3 X, 22 October 1970, K. Hickman (FSCA). Sumter Co., Oxford, 6 X, juv. X, 11 October 1971, L. Hubricht (VMNH). Louisiana: Calcasieu Par., Lake Charles, Duggan’s Greenhouse, 3 X, 2 August 1964, M. Kordish (FSCA). East Baton Rouge Par., Baton Rouge, Louisiana State University campus, nr Audubon Hall, earthworm culture, juv., 27 September 1964, R. E. Tandy (FSCA). New State Record. Illinois: Champaign Co., Urbana, in greenhouse, 3 X, 5 April 1953, collector unknown (INHS). Texas: Jim Wells Co., Alice, Green Thumb Nursery, 4 X, 22 September 1961, R. O. Albert (FSCA). New State Record.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF8DFF8DFD45CF4EFDFAAEA6.taxon	diagnosis	Diagnosis. Same as that of P. digitata but with 20 segments including epiproct, instead of 19. Variation. Not observed, with only one adult specimen and one juvenile known from the USA. Ecology. The two US specimens were encountered in a ‘ pine – oak association’ (Shelley and Golovatch, 2001). Overall distribution. Poratia obliterata is a neotropical species. The type locality is near Iquitos, Loreto Prov., Peru (Kraus, 1960), and the species is also known from Amazonia, Brazil, Costa Rica and greenhouses in Paris, France, and Germany (Adis et al., 2000; Golovatch and Sierwald, 2001). Occurrence in the USA. Known only from a single site in coastal Georgia representing an introduction (figure 1) (Golovatch and Sierwald, 2001; Golovatch et al., 2001). Published record. Georgia: Glynn Co., nr Brunswick, St. Simons I., rd to Fort Frederica (Shelley and Golovatch, 2001). New records. None.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF8DFF8FFD9AC993FC1FAF09.taxon	diagnosis	Diagnosis. Twenty segments. Preserved specimens thin, fragile and lightly sclerotized, colour whitish. Scalloped margin of collum upturned to varying degrees, often strongly; scallops moderately separated with deep indentations, lobes segregated to subequivalent degrees. Paranota 2 with three lobes, remaining paranota, both poriferous and non-poriferous, with two lobes; porosteles present on segments 5, 7, 9, 10, 12, 13 and 15, positioned between lobes (figures 4, 9). Variation. When viewed laterally, the strongly upturned collum of C. sanctus can lead to a misidentification as Poratia obliterata. The degree of upturn varies, but it is strong in most individuals and subequal to the condition in Poratia; the confusion is increased by the fact that the scallops are more distinctly separated than in Psochodesmus crescentis and the species of Myrmecodesmus, though less than in Poratia. The key to distinguishing C. sanctus from P. obliterata is the degree of segregation of the lateralmost scallops (compare figures 3 and 4); that to distinguishing C. sanctus from P. crescentis is the different paranotal configurations, with two lobes on all paranota of C. sanctus from segments 3 – 19 and porosteles between the lobes. In P. crescentis the non-poriferous paranota have three lobes with sometimes a faint suggestion of four lobes on segments 16 – 18, and there are two lobes anterior to the porosteles on segments 7 – 15 (compare figures 9 and 10). Ecology. The type of C. schubarti was found under a board (Causey, 1960), and the Alabama sample was in a ravine (Shelley and Golovatch, 2001). Habitat notations on vial labels include ‘ in soil’, ‘ in greenhouse’ and fire ant ‘ nest 76 - W- 048 A’. Overall distribution. The type locality is Pirassununga, São Paulo State, Brazil (Schubart, 1944), and C. sanctus also occurs in northwestern Argentina and Paraguay as well as the USA (Hoffman, 1993, 1999). However, the Argentina individuals have porosteles only on segment 5, a seemingly significant difference from the seven pairs on the introduced specimens in the USA, perhaps reflecting a different species. Occurrence in the USA. Eastern Alabama and the southern half of Georgia to Hillsborough and St. Lucie Counties, Florida (figures 1, 2). Hoffman (1999) stated that C. sanctus occurred nearly statewide in Florida, but the species is unknown from south Florida or the panhandle west of Leon County. Published records. Alabama: Tallapoosa Co. (Shelley and Golovatch, 2001). Florida: Alachua and Columbia Cos (Causey, 1960; Hoffman, 1993; Shelley and Golovatch, 2001); Clay, Duval, Hernando, Highlands, Hillsborough, Jefferson, Levy, Marion, Pasco, St. Lucie and Taylor Cos (Shelley and Golovatch, 2001). Georgia: Glynn, Grady and Turner Cos (Shelley and Golovatch, 2001). New records. Florida: Alachua Co., Gainesville, Hogtown Cr. Area, in soil, X, 10 October 1959, W. J. Platt III (FSCA) and Division of Plant Industry Bldg, 2 X, 1 July 1983, G. B. Edwards (FSCA); nr Melrose, 4 X, 8 November 1960, R. E. Woodruff (FSCA); and Paynes Prairie, on shoulder of US highway (hwy) 441, ‘ nest 76 - W- 048 A, ’ XX, 24 March 1976, Atwood, Hicks, Jouvenaz (VMNH). Citrus Co., Crystal River State Historic Site, 4 X, 11 June 1983, R. M. Shelley, J. L. Staton (NCSM). Hillsborough Co., N. Tampa, X, 30 March 1967, P. C. Drummond (FSCA). Jefferson Co., Big Bend Horticultural Lab. nr Monticello, 2 X, date and collector unknown (FSCA). Lake Co., Astor, three juvs., 12 September 1959, W. Suter (FSCA). Leon Co., Tall Timbers Res. Sta., X, August 1973, N. A. Naves (FSCA). Georgia: Chatham Co., Halycon Bluff, X, 1 November 1959, L. Hubricht (FSCA). Alabama: Crenshaw Co., Brantley, 4 X, 17 July 1960, L. Hubricht (FSCA). Dale Co., Ozark, 2 X, 25 June 1960, L. Hubricht (VMNH).	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF8FFF89FDAEC823FB64ADF6.taxon	description	figure 27 a, b; Chamberlin, 1951: 29; Chamberlin and Hoffman, 1958: 77; Hoffman, 1999: 497; Shelley and Golovatch, 2001: 62; Shelley, 2001: 247. Diagnosis. Nineteen segments. Preserved specimens thin, fragile and lightly sclerotized, colour tawny. Scalloped margin of collum essentially flat, not or only very slightly upturned; scallops shallowly and subequally separated. Paranota 2 with three lobes, remaining non-poriferous paranota generally with three lobes, caudalmost lobe on 16, 17 and / or 18 sometimes faintly indented to form four barely perceptible lobes; porosteles present on segments 5, 7, 9, 10, 12, 13 and 15, located at caudolateral paranotal corners, with one anterior lobe on segment 5 and two thereafter (figures 5, 10). Variation. The scallops on the collum are always shallowly segregated, and the condition of the paranotal lobes through segments 15 – 16 does not vary. However, the paranota on segments 16 – 18 show either three lobes or four very faint ones, based on a very slight indentation of the caudalmost lobe to form two barely perceptible sublobes. More often, there are just three lobes. The paranotal configuration, with porosteles at the caudolateral corners, is similar to that of species of Poratia, except there are two lobes anterior to the porosteles in Psochodesmus crescentis (only one on segment 5) and one anterior lobe throughout the body in Poratia. Ecology. Habitat notations with specimens examined during this study include ‘ floodplain of stream in hammock’, ‘ Pinus clausa litter’, ‘ mixed woods’, ‘ hardwood litter’, ‘ deep soil extraction, hardwood forest’, ‘ oak palm forest’, ‘ Eciton sp. nest’, ‘ litter of Pinus sp. ’ and ‘ sand pine litter’. Overall distribution. Psochodesmus crescentis is known only from peninsular Florida, where it may be endemic. Occurrence in the USA. The type locality is Crescent City, Putnam County, and P. crescentis is also known from between Crescent City and Palatka; Vero Beach, Indian River County; and Coconut Grove, Dade County (Loomis, 1934). Causey (1960) gave the range as ‘ Putnam Co. to the end of the peninsula’, and Hoffman (1999) recorded it as the southern two-thirds of the Florida peninsula. Causey’s characterization is reasonably accurate, as the species actually extends from Clay County (just north of Putnam) to the southern tip of the peninsula (figure 2). Published records. Florida: Dade, Highlands, Indian River and Putnam Cos (Cook, 1896; Loomis, 1934; Chamberlin, 1951). New records. Florida: Alachua Co., Magnesia Springs, ca 4 mi (6.4 km) W Hawthorne, juv. X, 21 September 1963, F. W. Mead (FSCA). Brevard Co., Merritt I., Kabboord Sanctuary, 1.5 mi (2.4 km) E FL hwys 3 and 520, XX, juvs, 23 February 1996, G. Steck, P. Skelley and Sutton (FSCA). Broward Co., Deerfield Beach, 2 X, February 1959, and X, 20 October 1959, H. A. Denmark (FSCA). Clay Co., 5 mi (8 km) W Green Cove Springs, fragment, 7 June 1928, collector unknown (NMNH); and 1.5 mi (2.4 km) NE Middleburg, 4 X, 8 August 1958, L. Hubricht (VMNH). Dade Co., Miami, Matheson Hammock, 3 X, 4 July 1977, R. M. Shelley (NCSM), XX, 9 July 1986, Klimaszewski, S. Peck (NCSM), X, 27 August 1986, S. and J. Peck (NCSM), and W, 9 December 1986, S. and J. Peck (NCSM); S. Miami, 7900 SW 176 th St., Old Cutler Hammock Park, X, 15 November 1985, S. and J. Peck (NCSM); and Everglades National Park, Royal Palm Hammock, X, 9 July 1986, S. and J. Peck (NCSM). DeSoto / Highlands / Okeechobee Cos, along FL hwy 70 between Arcadia and Okeechobee, 4 X, 10 April 1973, H. F. Loomis (VMNH) [plotted in DeSoto Co. in figure 2]. Hernando Co., along FL hwy S- 595, 5 mi (8 km) S jct. FL hwy 50, 10 X, 13 March 1966, P. C. Drummond (FSCA); and Brooksville, XX, 7 November 1929, H. F. Loomis (VMNH). Highlands Co., Highlands Hammock St. Park, XX, 15 June 1955, H. S. Dybas (FSCA) and WW, XX, 1 June 1958, N. B. Causey (FSCA); and Lake Placid, W, 2 X, 7 April 1950, D. E. Beck (NMNH) and Archbold Biological Station, W, 1958, N. B. Causey (FSCA). Hillsborough Co., Tampa, XX, March 1898, collector unknown (NMNH). Indian River Co., Vero Beach, X, 22 April 1933, H. F. Loomis (NMNH). Lake Co., between Astor and Astor Park, X, juv., 10 October 1929, collector unknown (NMNH). Marion Co., 10 mi (16 km) N Ocala, 5 X, juvs, March 1931, O. F. Cook (NMNH); 1.5 mi (2.4 km) S Moss Bluff, W, 22 May 1958, M. H. Muma (FSCA); and Juniper Springs, juv., 21 December 1959, H. A. Denmark (FSCA). Pasco Co., Dade City, XX, juvs, 8 December 1928, O. F. Cook (NMNH). Putnam Co., locality unknown, 2 X, 5 March 1950, A. Van Pelt (FSCA); and Welaka, University of Florida Reserve, 2 X, 8 April 1964, H. A. Denmark (FSCA). Seminole Co., 4 mi (6.4 km) NE Oviedo, 4 X, 10 November 1962, M. H. Muma (FSCA). Volusia Co., DeLeon Springs, XX, 10 October 1929, H. F. Loomis (VMNH) and X, 12 June 1983, R. M. Shelley and J. L. Staton (NCSM). Remarks. Loomis (1961: 100, figure 4 a) published a line drawing of the left gonopod of P. crescentis without an accompanying taxonomic account. Causey (1960) stated that P. crescentis is the ‘ most abundant micropolydesmoid on the Florida peninsula’ and is ‘ well established in rural areas’; it should be classified as a ‘ Florida endemic or a species known only from this state’, according to the criteria of Shelley (2001). The fragment from west of Green Cove Springs, Clay County, consists of only a few anterior segments, not enough to determine sex (not including segment 7) but enough to identify it positively as P. crescentis by the configurations of the paranota on both poriferous and non-poriferous segments.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF88FF8AFD81CCE1FC60A932.taxon	description	Myrmecodesmus formicarius Silvestri, 1910: 360, figure 5; 1911: 193, figures 15, 16; Attems, 1940: 312, figures 447 – 450; Loomis, 1968: 50; Hoffman, 1973: 513 – 515, figures 1, 2; 1999: 493; Shear, 1977: 254; Shelley and Golovatch, 2001: 60, figures 1, 2; Shelley, 2001: 247.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF88FF8AFD81CCE1FC60A932.taxon	diagnosis	Diagnosis. Twenty segments. Preserved specimens hard, heavily sclerotized, usually dark brown, grey, or black in colour. Scalloped margin of collum not or only very slightly upturned; scallops very shallowly and equivalently separated. Paranota 2 with three lobes, remaining paranota with two lobes; poriferous paranota with porosteles positioned between lobes, anterior lobe much larger than caudal. Porosteles present on segments 5, 7, 10, 13 and 15. Paramedian tubercles fused into prominent crests; those on caudal segments narrowly and shallowly segregated, overhanging and extending beyond level of caudal extremity of epiproct, concealing latter in dorsal view (figures 6, 13, 14). Variation. The degree of indentation of the scallops on the collum varies considerably, from the slightly indented lobes illustrated by Shelley and Golovatch (2001: figures 1, 2) to deep indentations that are subsimilar to the condition in C. sanctus. Ecology. The types of I. cajuni were collected under the bark of orange trees. The sample from Cameron County, Texas, was encountered ‘ under palm logs’, but most of the new samples from Texas were taken in caves. Overall distribution. The type locality, the southernmost known site, is Xalapa, Veracruz, Mexico (Silvestri, 1910; Attems, 1940; Loomis, 1968; Hoffman, 1973), and M. formicarius ranges northward into northern Texas (figure 1). It is also known from southeastern Louisiana, an area including the type locality of the synonym, I. cajuni, Venice, Plaquemines Parish. Occurrence in the USA. Myrmecodesmus formicarius occurs from the Rio Grande Valley in the southern tip of Texas northward to Dallas County [a distance of some 445 mi (712 km)] and westward to Sutton County. This coherent distribution, extending northward from Veracruz, appears to be natural, and the species appears to be indigenous to this area. It is also known from Pointe Coupee, Plaquemines and Orleans parishes, Louisiana, a detached area that I believe results from human importations. The record of M. digitatus from Calhoun County, Florida (Hoffman, 1999), that was assigned to M. formicarius by Shelley and Golovatch (2001), is returned to M. digitatus, as a direct comparison with the holotype showed it to be unequivocally this species. Myrmecodesmus formicarius is therefore deleted from Florida, where it was listed as an inhabitant by Shelley (2001). Published records. Texas: Cameron, Goliad and Guadalupe Cos (Loomis, 1959; Hoffman, 1973; Shelley and Golovatch, 2001). Louisiana: Orleans, Plaquemines and Pointe Coupee Pars (Loomis, 1944; Hoffman, 1973, 1999; Shelley and Golovatch, 2001). New records. Louisiana: Plaquemines Par., Boothville, X, 22 January 1950, G. H. B. (FSCA). Texas: Bexar Co., Camp Bullis, Lone Gunman Pit, X, 23 October 1997, P. Sprouse and G. Veni (TMM). Burnet Co., Simon Says Sink, 12 mi (19.2 km) NE Burnet, X, 12 November 1990, J. R. Reddell and M. Reyes (TMM); and Marble Falls, Railroad Cave (Cv.), X, 5 September 1993, A. G. Grubbs, T. Whitfield and G. Hoese (TMM). Caldwell Co., Maxwell, WW, XX, 1 December 1963, R. O. Albert (FSCA). Cameron Co., near Southmost, Rabb Ranch, WW, XX, 12 December 1954, L. Hubricht (VMNH). Comal Co., Camp Bullis, Camp Bullis Cv. No. 1, X, 25 January 2000, J. R. Reddell and M. Reyes (TMM). Dallas Co., DeSoto, off FM 1382, Whitewater Trail, X, 17 March 2002, C. T. McAllister (NCSM). Goliad Co., Goliad St. Park, X, 7 April 1954, L. Hubricht (VMNH). Guadalupe Co., Sequin, near Guadalupe R., W, X, 4 June 1955, L. Hubricht (VMNH). Hays Co., Ladder Cv., 11 mi (17.6 km) W San Marcos, juv., 26 May 1989, A. Grubbs and J. R. Reddell (TMM). Kendall Co., ‘ The Crack, ’ X, 28 May 1990, D. Pate (TMM). Kerr Co., Kerrville, WW, XX, 20 April 1962, R. O. Albert (FSCA); and 5 mi (8 km) S Kerrville on Guadalupe R., W, X, 20 April 1962, R. O. Albert (FSCA). Sutton Co., Caverns of Sonora, upper rest area, X, 9 December 1997, G. Veni (TMM). Travis Co., Austin, 9 mi (14.4 km) SSE State Capitol, Goat Cv., X, 23 January 1991, J. R. Reddell and M. Reyes (TMM), 10 mi (16 km) SSE State Capitol, Midnight Cv., X, 29 October 1999, M. Sanders (TMM), 12 mi (19.2 km) NW State Capitol, Wooden Derrick Cv., X, 14 April 1991, W. Elliott (TMM), 12 mi (19.2 km) NW State Capitol, Rockpile Cv., 3 X, 3 February 1989, J. R. Reddell and M. Reyes (TMM), and 13 mi (20.8 km) NW State Capitol, Cactus Pit, X, 19 June 1997, J. R. Reddell and M. Reyes (TMM); and along Buttercup Cr., Karst Area, Windmill Cv., X, 26 September 2000, J. R. Reddell, M. Reyes and M. Warton (TMM). Williamson Co., 2 mi (3.2 km) S Georgetown, Inner Space Caverns, X, 22 December 1968 to 5 January 1969, W. Elliott (TMM); Sun City, 10 mi (16 km) NW Georgetown, Do Drop In Cv., W, 28 July 1995, W. Elliott (TMM) and 12 mi (14.4 km) NW Georgetown, Good Omen Cv., X, 30 July 1994, J. R. Reddell and M. Reyes (TMM); Georgetown, S Bluff of N. San Gabriel R., Paleospring Cv., 2 X, 28 July and 2 August 1994, P. Sprouse and B. Larsen (TMM); 3 mi (4.8 km) W Round Rock, Buck Pride Cv., W, 4 X, 21 and 29 May 1996, J. R. Reddell and M. Reyes (TMM); Lakeline Cv., 1 mi (1.6 km) W jct. FM 620 and US hwy 183, X, December 1985, D. L. Pate (TMM) and X, 19 October 1990, J. R. Reddell and M. Reyes (TMM); Jug Cv., 2 mi (3.2 km) N jct. FM 620 and US hwy 183, X, 14 March 2000, J. R. Reddell, M. Reyes, P. Sprouse and G. Veni (TMM); and Raccoon Cv., 1 mi (1.6 km) N jct. US hwy 183 and FM 620, X, 27 March 1990, J. R. Reddell and M. Reyes (TMM). Remarks. The distributions of M. formicarius and digitatus in the USA parallel to some degree those of the two species of Desmonus Cook (Polydesmida: Sphaeriodesmidae), which are unquestionably indigenous. Desmonus pudicus (Bollman) ranges from Nuevo León, Mexico, northward through Texas to southwestern Missouri, and D. earlei Cook occurs along the Gulf Coast from eastern Mississippi to northern Florida, extending northward to southwestern Virginia and central Kentucky (Shelley, 2000). The distributions of M. formicarius and digitatus, though less extensive, demonstrate this same pattern, suggesting that most of their occurrences are native as well. One species, M. formicarius, ranges northward from Mexico a substantial distance into the USA, and the other, M. digitatus, occurs along the Gulf Coast. I place M. colotlipa in synonymy based on concordance in somatic features, particularly the crests on the caudal segments, which exhibit the configuration illustrated in figures 13, 14. The type locality is Cueva de Jutxlahuaca, at or near Colotlipa, Guerrero, Mexico, some 190 mi (304 km) WSW of Xalapa, the type locality of M. formicarius. The type collection (NMNH) houses two vials of I. colotlipa, one containing three females and juvenile ‘ types’, collected in this cave by F. Bonet on 16 May 1941, and the other containing a male and an unknown number of female ‘ paratypes’, all highly fragmented, taken there by the same collector on 16 January 1941. The latter sample is so highly fragmented that I could not even determine the total number of individuals, but one piece does contain a gonopodal aperture, so there was at least one male. The gonopods are not in the aperture and are lost, so new collections are needed to compare their structures against published drawings of M. formicarius (Hoffman, 1973: figures 1, 2). However, the virtually identical somatic features and the fact that the native occurrence of M. formicarius ranges much farther northward from Xalapa than the distance westward to Guerrero lead me to propose this synonymy.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF8AFF8AFD88CE11FB46ADF9.taxon	diagnosis	Diagnosis. Same as for M. formicarius except crests on caudal segments widely and deeply segregated, usually diverging slightly caudad, overhanging epiproct, tip of latter visible in dorsal view between crests (figures 15, 16). Variation. Contrary to what Shelley and Golovatch (2001) said, the difference in the lengths of the caudal crests and the degree to which they extend beyond the epiprocts in M. formicarius and digitatus (figures 14, 16) is too insignificant to base determinations on it. The only distinguishing somatic feature involves the degree of segregation of the caudalmost crests and the fact that the epiproct is always visible between them in M. digitatus and never so in M. formicarius (figures 13, 15). The caudal crests typically diverge slightly in M. digitatus but occasionally extend linearly in a subparallel arrangement. Ecology. The types were collected from beneath logs; the sample from Hancock County, Mississippi (Shelley and Golovatch, 2001) was found ‘ lakeshore along railroad’. Habitat information on vial labels examined during this study include ‘ in sugarcane’, ‘ under palm logs’ and ‘ 1 in. deep in soil of fungus garden of Atta texana ’. Overall distribution. The eastern panhandle of Florida to north-central Alabama and western and northern Louisiana (figures 1, 2); the type locality is along US highway 190 between Kinder and LeBlanc, Allen Parish, Louisiana. Myrmecodesmus digitatus has never been reported outside this area and is thus both indigenous and endemic to the Gulf Coastal Plain. Occurrence in the USA. Same as the overall distribution. Published records. Florida: Calhoun Co. (Hoffman, 1999). Mississippi: Hancock Co. (Shelley and Golovatch, 2001). Louisiana: Allen Par. (Loomis, 1959; Causey, 1963; Shelley and Golovatch, 2001). New records. Florida: Calhoun Co., Scotts Ferry near Chipola R., X, 6 September 1959, L. Hubricht (VMNH). This is the record by Hoffman (1999), which was not provided in detail and was erroneously assigned to M. formicarius by Shelley and Golovatch (2001). Alabama: Baldwin Co., W of Loxley, jct. US hwys 90 and 98, X, 2 July 1960, collector unknown (FSCA). Jefferson Co., Homewood, X, 25 April 1960, H. R.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF94FF97FD87CD0BFBE9A9D5.taxon	materials_examined	Type specimens. One male and three female syntypes (FSCA) collected by J. R. Reddell, 15 March 1969, 7.5 mi (12.0 km) N Boerne, Kendall County, Texas. I designate syntypes instead of a holotype and paratypes because, after noting somatic features of the male, I had to dissect the gonopods, which resulted in fragmentation such that I could no longer distinguish the parts of the male from those of the already fragmented females. Diagnosis. Porosteles present on segments 5, 10, 13 and 16. Crests on caudalmost segments diminished in height and only slightly overhanging sides of epiproct, barely extending beyond level of its caudal extremity, tip of epiproct visible in dorsal view between crests (figures 17, 18). Male syntype. Twenty segments including epiproct. Length ca 3.4 mm, maximum width ca 0.6 mm. Colour subuniformly dark brown; body heavily sclerotized. Collum with low, rounded indistinct pustules, anterior margin relatively flat, not appreciably upturned, scallops moderately distinct, middle two most strongly segregated, lateralmost shallowly and indistinctly separated (figure 7). Second tergite with typical three-lobed paranota, paramedian tubercles coalesced into low, indistinct carinae, caudal segmental margin smooth and sublinear, slightly crenulated laterad. Remaining tergites more strongly crenulated laterad, carinae becoming progressively more distinct through segment 17 though always lower than in M. formicarius and M. digitatus; carinae on segments 18 – 19 reduced, those on 19 slightly overhanging sides of epiproct, tip of latter visible between crests (figures 17, 18). Paranota with two moderately distinct lobes on all segments subsequent to segment 2, lobes subequal in size on non-poriferous segments, anterior lobe much larger on poriferous segments; porosteles present on segments 5, 10, 13 and 16, positioned between lobes (figure 11). Epiproct small and smoothly rounded, without lobes or crenulations. Legs and sterna without modifications. Gonopod (figures 19, 20) minute, details visible only under high magnification; telopodite generally uncinate, presumed solenomere a short, gently curved projection arising at two-thirds length on caudal surface, tip of telopodite divided into three branches, anteriormost subdivided. Prefemoral process much shorter than telopodite, bent caudad distally, apically bifurcate. Female syntypes. The female syntypes are light brown in colour but otherwise agree closely with the male in somatic features. Ecology. The habitat at the type locality is not indicated on the vial label. Distribution. Known only from the type locality. With only one sample, we have no way of determining whether M. reddelli is introduced from a neotropical locale, native to this region of Texas, or possibly even endemic. I suspect that, like M. formicarius, it is a Mexican species that ranges northward into southern Texas and hence is indigenous. Etymology. I am pleased to name this species for the collector, J. R. Reddell, of the Texas Memorial Museum, Austin, in recognition of his continued helpfulness to me in providing research material and his lifetime of collecting myriapods in Texas and Mexico, particularly from caves. Remarks. In studying this milliped, I examined holotypes of every species from Mexico and one from Cuba that appeared from published descriptions to have a somewhat similar gonopod (listed in the acknowledgements), but none matched M. reddelli either in somatic features, gonopodal configuration or both; the closest was M. mundus (Chamberlin), from Veracruz (city), Mexico, some 800 mi (1280 km) to the south. Even at 400 ×, the highest magnification on my Nikon compound microscope, I was unable to resolve the course of the prostatic groove and which branch is the solenomere, which may require scanning electron microscopy. However, a review of the literature (Shear, 1973, 1977) indicates that it is probably the proximal branch as in other species of the genus, and Dr Shear (in litt.) has so advised me, hence the characterization of ‘ presumed solenomere’ in the description and the label as such on figures 19, 20. As M. formicarius ranges northward from Mexico into Texas, it seems likely that M. reddelli does too and that it is indigenous, but not endemic, to Texas. Without question, there are many undiscovered pyrgodesmid species in Mexico and Central America, and there may be still others that range northward across the Rio Grande into the southern reaches of Texas.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
03EB87D5FF97FF91FD87CF47FB47AE57.taxon	description	Lophodesmus bituberculatus Loomis, 1970: 130 – 131, figure 1; Hoffman, 1999: 489. Lophodesmus caraibianus: Shelley, 2001: 246 – 247. Diagnosis. Twenty segments. Preserved specimens hard and heavily sclerotized, usually dark grey in colour. Scalloped margin of collum essentially flat, not or only very slightly upturned; scallops very shallowly and equivalently separated. Paranota 2 with three lobes, remaining non-poriferous paranota with two lobes; porosteles present on segments 5, 7, 9, 10, 12, 13, 15 and 16, positioned at caudolateral corners, with one, broad, subtruncated anterior lobe (figures 8, 12). Variation. Process ‘ B’ (figure 21) is less distinct on the holotype gonopod. Ecology. Causey (1960) was the first to report ‘ Lophodesmus ’ from Florida, which she misidentified as ‘ L. ’ caraibianus (Chamberlin); her record was based on material collected under a board. Notations on vial labels of material examined during this study include: ‘ berlese of oak hammock litter’, and ‘ deep soil extraction, hardwood forest’. Overall distribution. Known previously only from the type locality, Cueva Pajita, Lares, Puerto Rico. Distribution in the USA. Known only from Dade County, Florida (figure 2). Published records. Florida: Dade Co. (Causey, 1960; Shelley, 2001; both as L. caraibianus). New records. Florida: Dade Co., Miami, 2 X, 17 June 1959, R. E. Woodruff (FSCA, VMNH); Matheson Hammock, 2 X, 4 July 1977, R. M. Shelley (NCSM) and W, three juvs, 9 July to 9 December 1986, S. Klimaszewski and J. Peck (NCSM); South Miami, 7900 SW 176 th St., Old Cutler Hammock, 5 W, XX, 15 November 1985, S. Peck (NCSM) and juv., 26 August 1986, S. and J. Peck (NCSM); and Everglades National Park, Royal Palm Hammock, two juvs, 9 December 1986, S. and J. Peck (NCSM). New State Record. Remarks. In determining the identity of this species, direct comparison with the type immediately revealed that it is not ‘ L. ’ caraibianus (Chamberlin), which is a larger-bodied species with different dorsal sculpturation. Causey (1960) did not indicate the basis for her determination of the specimens collected in 1959, and I (Shelley, 2001) repeated the record. Chamberlin (1918) did not note the sex of his two specimens of ‘ L. ’ caraibianus in the original description nor does Loomis (1934, 1936), but Hoffman (1999) reported that the one then remaining was a female. It is actually a male with the coxae present in the aperture, but the telopodites have been broken off and are lost. Loomis (1936) reported males from four sites in Haiti and briefly characterized the gonopods but did not illustrate them, and it is unknown whether this material is conspecific with the holotype, collected at Mandeville. ‘ Lophodesmus ’ is beset with formidable taxonomic problems, as detailed by Hoffman (1976, 1999), and the uncertain identity of ‘ L. ’ caraibianus should be added to these difficulties. The type species, L. pusillus Pocock, occurs in Indonesia (Flores), and the New World species likely are not congeneric (Hoffman, 1999). I therefore place the generic name in quotation marks herein, except when citing literature accounts per se. In checking the descriptions and illustrations of the known species of ‘ Lophodesmus ’ in the New World, I noted that Loomis’ (1970) gonopod drawing of the type of ‘ L. ’ bituberculatus shows two short, pointed projections on either side of the telopodite, suggesting the structures labelled ‘ A’ and ‘ B’ in figure 21 of a male from Miami. No other published drawing of a ‘ Lophodesmus ’ gonopod shows such features, but as it was drawn at low magnfication, it does not begin to show the complexity of the telopodite. I therefore compared the gonopods of a Miami male against those of the holotype to confirm the determination. The list of specimens examined includes all the known material of ‘ L. ’ bituberculatus from the continental USA. It is noteworthy that H. F. Loomis never collected the milliped in Dade County despite living in South Miami from the 1950 s until his death in 1976. The first collection was by R. E. Woodruff in 1959, then by me in 1977, and subsequently four times by Stewart and Jarmila Peck. Consequently, the introduction of ‘ L. ’ bituberculatus into south Florida seems to have occurred quite recently, and substantial populations may have developed. This exogenous, neotropical milliped may soon spread into adjoining parts of Broward and Monroe Counties.	en	Shelley, R. M. (2004): The milliped family Pyrgodesmidae in the continental USA, with the first record of Poratia digitata (Porat) from the Bahamas (Diplopoda: Polydesmida). Journal of Natural History 38: 1159-1181
