THE MILLIPEDE GENUS CYLINDROIULUS VERHOEFF, 1894 (DIPLOPODA: JULIDA; JULIDAE) IN SPAIN: RECENT RECORDS AND DESCRIPTION OF THREE NEW SPECIES FROM THE BASQUE COUNTRY, NAVARRA, ANDALUCÍA AND CATALUÑA; El género de milpiés Cylindroiulus Verhoeff, 1894 (Diplopoda: Julida; Julidae) en España: citas recientes y descripción de tres nuevas especies del País Vasco, Navarra, Andalucía y Cataluña

Two large collections of Cylindroiulus were examined (one from the Basque country and Navarra and one from Andalucía) supplemented by other, smaller collections from other areas of Spain; in total 10 species were represented. Three new species are described: C. caramelos sp. nov., C. elosegiorum sp. nov. and C. karinhansorum sp. nov. The remaining seven species are: C. londinensis (Leach, 1814), C. franzi Attems, 1952, C. punctatus (Leach, 1815), C. pyrenaicus Brölemann, 1897, C. sagittarius Brölemann, 1897, C. sanctimichaelis Attems, 1927 and C. unciger Attems, 1952. Information has been presented on these species where this has added to that previously available. Comments have been made on the challenging londinensis-group which seems particularly specious in the Iberian Peninsula. Suggestions are made for further work on this group include genetic studies.


Introduction
The diplopod fauna of the Iberian Peninsula is relatively under studied in comparison with much of Europe. Following work by Attems and Verhoeff in the early part of the 20 th century there was a relatively long gap until work in the 1970s and 1980s on cave fauna (Mauriès & Vicente, 1977a), species from northwest Spain and the Pyrenees (Mauriès, 1975;Mauriès & Vicente, 1977b) and some ecological studies by Vicente & Ascaso (1990) and Serra et al. (1997) in Cataluña and Mauriès & Barraqueta (1985) in Biscay. Vicente (1981) published a list of species from Cataluña including details to help with identification.
The French fauna is comparatively well studied and some Cylindroiulus species relatively abundant further north have a range that extends down to and/or over to the south side of the Pyrenees, such as Cylindroiulus londinensis (Leach, 1814), Cylindroiulus parisorum (Brölemann & Verhoeff, 1896) and Cylindroiulus punctatus (Leach, 1815) (Kime & Enghoff, 2017).

sysTemaTIcs of The genus Cylindroiulus
The genus Cylindroiulus is large, with over 100 species described. Read (1990) clarified a previously confusing situation between Cylindroiulus and Allajulus C.L. Koch, 1847 which are now considered separate genera, distinguished by the presence of setae on the apodous body rings in Allajulus and details of the gonopods. Both genera are found on the Iberian Peninsula, although Allajulus is currently represented only by A. spinosus (Ribaut, 1904). A few species groups can be identified within the genus, with those relevant to Iberia being: the perforatusgroup (Cylindroiulus perforatus, (Vehoeff, 1905) Cylindroiulus fenestratus, Read, 1989, Cylindroiulus ventanaea Read, 2007 and Cylindroiulus villumi Reboleira & Enghoff, 2018), distinguished by usually having a perforate mesomerite, a feature which has so far not been seen on any species outside the Iberian peninsular, and denticles on the anterior flagellumconducting lamella (Read, 1989;Reboleira & Enghoff, 2018), a feature which has so far not been seen on any species outside the Iberian Peninsula; some of the truncorum/luscus-group (arborum Verhoeff, 1928, bellus (Lignau, 1903, britanicus (Verhoeff, 1892), latestriatus (Curtis, 1845), parisiorum, and

DIagnosIs
Very similar to Cylindroiulus sanctimichaelis and Cylindroiulus caeruleocinctus, with large body size and lacking a telson projection. Differing by having a relatively long body with many body rings. Dark in colour, preserved specimens are banded in colour, with pale legs. Gonopods are similar to C. caeruleocinctus/C. sanctimichaelis, the anterior gonopods are relatively large in relation to the posterior gonopods which have a large brachite.

DescrIPTIon
Length: ♂︎ 38 mm; ♀︎ up to 51 mm. Body height: ♂︎ 2.1 mm; ♀︎ up to 3.06 mm. Rings: ♂︎ 57; ♀︎ up to 62. Body colour: General appearance, black with white legs . Head dark brown, prozonites pale grey, metazonites black, with paler posterior edge. characters that are generally useful for distinguishing species, making descriptions and identification of the species within the genus difficult. In the following descriptions the overall body shape is described using the terms anterior constriction, which is the reduction in body width behind the head in dorsal view. Posterior attenuation refers to the degree of reduction in body height at the posterior end of the body when looking from the side. The term body rings is used instead of segments as the rings in a millipede do not equate to true segments. Setae on the anal valves are generally evenly spaced and located just in from the margin, unless otherwise specified.
Locations given on labels were converted to a standard format (decimal degree) using coordinatesconverter.com. For the small number where details of location from labels were vague, a map was used to find the decimal degree of the nearest place mentioned on the label. To produce the distribution maps Henrik Enghoff hand plotted them on the relevant species map from the European Atlas (Kime & Enghoff 2017) using https://mappingsupport.com/p2/gissurfer.php to help locate the positions.

lIfe hIsTory
Adults were found in April. There are not enough specimens to analyse for life history.

DIsTrIbuTIon
Only known from the type locality (Fig. 18). When the manuscript was in the proof stage, communication with Hans Reip revealed that this species may be quite common west and south of the Ebro delta, a little south of the type location. Of the few that appeared unbroken, the accessory claws seemed to be slightly shorter than the claw (Fig. 6).
Females also had a large number of broken accessory claws, with those intact appearing to be slightly shorter than the claw. Telson: Preanal ring with no projection, perhaps bearing setae but probably worn off in the male. Subanal scale not projecting, with one pair of setae. Anal valves with three pairs of setae. Mature male secondary sexual characters: Mandibular stipites of mature males strongly expanded and projecting anteriorly, First pair of legs in mature male tight hooks with prominent 'knees' on the angle of the hook.
Gonopods : Promerite narrow and slender in anterior view, relatively bulky in mesal view, with hooked apex. Mesomerite narrow and slender and almost completely enclosed by promerite apart from the distal part. Flat topped and with different textured apex. Opisthomerite broad and bulky in mesal view. Brachite large, projecting beyond the solenomerite, with angular ventral margin and clear cleft between it and the solenomerite with membranes across it. Phylacum large, curvaceous. Solenomerite with three small teeth at the apex and membranous area at tip. Paracoxal rim very high, paracoxal process lacking.

eTymology
Named for its superficial resemblance to a mint humbug (a stripy child's sweet); "caramelos" are sweets in Spanish.

relaTIonshIPs wITh oTher sPecIes
This species clearly falls into the problematic londinensis group, for full discussion see under C. sanctimichaelis. It can be distinguished by details of the gonopods and has more body rings in total and more body rings for the body height than other similar species . The short accessory claw might also be diagnostic but most appeared to be broken so the reliability of this character is uncertain. It is also not well enough studied in all the similar species to be confident of its value. habITaT The type locality is situated in a limestone mountainous region range close to Tarragona with a Mediterranean Figs. 16-18.-16-17 16-18.-16-17. Gráficos de dispersión de la altura corporal frente a los anillos corporales con patas para C. sanctimichaelis y C. caramelos sp. nov. Ver el texto del apartado de C. sanctimichaelis para más detalles. 16. Machos. 17. Hembras (téngase en cuenta que no había hembras de Guadalajara). 18. Mapa de distribución que muestra la ubicación del tipo de C. caramelos sp. nov. being C. apeninnorum (Brölemann, 1897) (Italy, also presumably introduced to the Netherlands and UK), C. gestri (Silvestri, 1898) (Sardinia), C. propinquus (Portugal) and C. pyrenaicus (Spain). The gonopods however do not appear similar to any of these species.
It is unknown what the function of a strongly pointed telson is, and this is also true of the subanal scale.
Pointed telsons occur in a variety of Cylindroiulus species and it seems likely that this does not relate to any particular evolutionary relationships.
lIfe hIsTory Cylindroiulus species add the apodous body rings to the number of podous body rings at each moult, and, if they have ommatidia they also add eye rows each time so it is possible to infer details of their life history by looking at the distribution of these characteristics across the collected specimens (Read, 1988). The eye rows for this species are very jumbled and are also reduced in terms of numbers of ommatidia (in comparison to the total, assuming that an extra row of ommatidia is added at each moult). As part of the characterisation of this species the numbers of ommatidia were counted and a plot showing the number of individuals for each ommatidia number, broken down by sex is shown in Figs. [35][36]. Sometimes the number of ommatidia was not easy to determine and for a few specimens both sides of the head were examined. When the number of ommatidia varied on each side the head the lowest number was used in the graphs. Similar graphs illustrate the variation in the number of podous rings broken down by sex . These figures show that the mature males have a relatively tight cluster of possible podous ring numbers. Allowing for the relatively high number of apodous rings in this species, between 3 and 5 for individuals with 39 or 40 podous rings, this probably reflects just one or two stadia. Aside from one very small mature male (perhaps misdesignated?) the distribution seems to show a normal pattern of juvenile males developing into matures for both number of ommatidia and podous ring number.
The graphs for the females show a more extended 'tail', i.e. more individuals with larger numbers of ommatidia and podous rings than the males. There is therefore a suggestion that the females undergo further moults once maturity has been reached. This situation is not uncommon in the genus Cylindroiulus. It therefore appears that the males are probably semelparous, but the females may be iteroparous. Having eight apodous rings was not unusual amongst juveniles of this species, hence the gaps in the graphs reflects this development pattern.

DIsTrIbuTIon
This species was found quite widely and frequently over the area sampled in 2009, in the provinces of Navarra and Guipuzkoa and just into Alava. A species of the hills, it was found in several mountain ranges visible, pale brown/orange. Overall body shape: Anterior constriction not obvious, posterior part of trunk wider than anterior. Head: Antennae: slender and pale in colour. Eye rows: Eyes not in clear rows except in younger juveniles, difficult to 'read'. Ommatidia reduced in number, often jumbled and sometimes varying in size in the same ocular field (Fig. 21). Ommatidia black and contrasting in comparison with pale body colour. ♂︎ with 8-14 ommatidia; F up to 15-16 ommatidia. Body rings: Striae on metazonites fairly strong and close together, limbus crenulate (Fig. 22). Legs: Generally pale, similar to body colour. With strong and crooked setae on more basal leg segments. Claw of normal shape; accessory claw slightly longer than claw, with distinct change of direction after surpassing the claw tip (Fig. 23). Telson: Preanal ring with strong projection, curved ventrally. With 1 pair of setae at end and another pair ventrally. Less strongly projecting in juvenile individuals and tending to be straighter/less down turned. Subanal scale projecting well past the anal valves (Figs. 20,24). More-or-less parallel with ventral margin of trunk but sometimes appearing dorsally directed giving the impression of a pincer shape at the posterior end of the animal. With one pair of setae. Smaller relative to body size in juveniles (Fig. 25). Anal valves: With three pairs of setae (Fig. 24).
Mature male secondary sexual characters: Mandibular stipites of mature males barely expanded (Fig. 21). First pair of legs in mature males forming tight hooks with prominent 'knee' (Fig. 26). Keel of body ring 7 in mature males open, but not very wide, no obvious keel (Fig. 27) Gonopods (Figs. 28-33): Promerite: Short and squat with no 'window'. Flagellum of normal length. Mesomerite: Short and squat, similar in length to promerite. Opisthomerite: Simple in shape with membranous tip. Long thin phylacum and small brachite. Large paracoxal process which seems a slightly strange shape from some angles and in posterior view.

eTymology
Named after the Elosegi family from Leitza, in particular José Miguel, Migel Mari and Arturo, all of whom have helped me explore the Basque area and given freely of their time and knowledge. Note that members of the family spell their name in two different ways, Elósegui and Elosegi. The latter is the 'original' Basque spelling and the former the Spanish version. The family have expressed their preference for the Basque spelling.

relaTIonshIPs wITh oTher sPecIes
This species is one of a small group that have strongly down pointed dorsal projections on the preanal ring and strongly pointed subanal scale (although it is perhaps the most extreme in this species), others  Figs 35-39.-35-38. Gráficos mostrando los espectros de estados para C. elosegiorum sp. nov. 35. Número de individuos frente a número de omatidios en machos. 36. Número de individuos frente a número de omatidios en hembras. 37. Número de individuos frente a número de anillos con patas en machos. 38. Número de individuos frente a número de anillos con patas en hembras. 39. Mapa de distribución que muestra las localidades de colecta de C. elosegiorum sp. nov. present. Opisthomerite simple in shape, with hyaline region apically. Paracoxal process large and with jagged posterior margin, although rather less stout than illustrated by Mauriès (1978). Despite the size of the paracoxal process it can be difficult to see in situ being situated on the mesal side of the gonopods.

relaTIonshIPs wITh oTher sPecIes
Attems considered that this species was in the subgenus Plagionicus, together with C. parisiorum and C. bouvieri (Brölemann, 1896) Geoffroy et al. in press), but the simple gonopods look more similar to the Caucasian species C. placidus (Lignau, 1903) and C. besucheti Strasser, 1975 both considered in the C. placidus-group (Read, 1992). The original illustrations are rather poor, but a more complete description has been made by Mauriès (1978) including a drawing of the type. The current specimens accord well with that of Mauriès loc cit.

ecology & lIfe hIsTory
With so few samples further analyses are not possible.

DIsTrIbuTIon
Previously known from Sierra de las Nieves close to Ronda and Mt Arastepa in the Sierra de Ronda. The current larger collection is from the same area. If the collection from Granada can be confirmed as the same species (see below) this would be a considerable expansion to the range, although still within Andalucía. (Fig. 46).

Cylindroiulus cf. franzi
No mature males were in these collections. The telson and overall appearance looks similar but location of one is somewhat distant from those recorded so far. from Leitza in the east, across Sierra de Aralar to Gorbeia in the west, and south to Sierra de Urbasa and Sierra de Andia (Fig. 39). While many locations were deciduous woodland, frequently with Fagus sylvatica L. (Fig. 34), it was also found in stony wood pasture with scattered trees and in the garden of a house in the centre of the town of Leitza. Attems, 1952 Figs. 40-46 Material studied:  (4), females (24) DIagnosIs A small and pale species with no telson projection but a characteristic tented shape to the top of the telson. Simple shaped gonopods with a prominent paracoxal process which is jagged on the posterior side.

DescrIPTIon
Length: ♂︎ 17 mm; ♀︎ up to 20 mm. Body height: ♂︎ 1 mm; ♀︎ up to 1.3 mm. Rings: ♂︎ 48-51, ♀︎♀︎ up to 54. Body colour: Overall pale in colour, telson a little darker, some individuals with small amount of darker mottling. Antennae and legs pale in colour. Overall body shape: Anterior constriction behind head; posterior attenuation only of apodous rings. Legs with accessory claw slightly longer than claw. Head: Eye rows rather jumbled with what look to be extra ommatidia (Fig. 40). First two rows often seemingly a single ocellus. Stadia: ♂︎ VII to VIII, ♀︎ up to IX. Body rings: Striae on metazonites regular. Limbus not crenulate. Telson: Preanal ring without clear dorsal projection but with a roof-like shape which gives the appearance of a prominent point dorsally (Fig. 41). Bearing 1 or 2 pairs of setae ventrally and one pair dorsally on the projection. Subanal scale not projecting, bearing 1 pair of setae. Anal valves generally with 3 pairs of setae.
Mature male secondary sexual characters: Mandibular stipites of mature males expanded to medium extent, with a vertical anterior margin (Fig. 40). First pair of legs in mature male: tight hooks. Keel of body ring 7 in mature male very small, barely projecting but body appearing slightly wider at this point.
Gonopods : Promerite and mesomerite relatively low and squat. Promerite slightly longer and much broader than mesomerite which appears narrowly projecting up almost through the promerite. Flagellum of normal length. Paracoxal rim barely Gonopods : Promerite slightly shorter than mesomerite, flagellum of normal length. Opisthomerite simple in shape, parallel sided and lacking brachite but with possible slight phylacum. Apex with hyaline structure. Paracoxal process sharp, clearly defined but not extensive. Paracoxal rim not particularly large. eTymology Named in honour of Karin Voigtländer and Hans Reip who collected the type material, contributed many other specimens to this study and who have both contributed greatly to our knowledge of millipedes.

relaTIonshIPs wITh oTher sPecIes
Cylindroiulus karinhansorum sp. nov. does not easily fit into a Cylindroiulus species group. The relatively simple gonopods, similar to the Cylindroiulus ground plan make it difficult to comment on its taxonomic position; they are not dissimilar to C. franzi in overall appearance.

ecology & lIfe hIsTory
Six females from the type locality had fungi on their first two pairs of legs ( Fig. 56) which was identified as Rickia laboulbenioides De Kesel by Henrik Enghoff. The species seems to have been found in wooded areas in the mountains, often with Quercus suber.
The stadial spectra of this species is given in Figs. 57-58. The eye rows were not clear enough to consider reliable, so the graphs have been drawn using the number of podous rings. Generally mature males had up to four apodous rings, the graph thus suggests that the mature males probably represent two different stadia, suggesting that they are semelparous. As expected, the females had a wider range of numbers of podous rings, with one having 59, substantially more than the others. There are hints that some females, at least might be iteroparous.

DIsTrIbuTIon
Discovered from several locations in the Grazalema mountains and nearby in P.N. de los Alcornocales (Fig. 59).

DIagnosIs
A medium sized species, generally pale in colour. With slightly, but definitely, projecting telson and relatively simple gonopods.

DescrIPTIon:
Length: ♂︎ 13 mm; ♀︎ up to 24 mm. Body height: ♂︎ 0.8-1.0 mm; ♀︎ up to 1.5 mm. Rings: ♂︎ 37-44+3-4 apodous (one male apparently has 49 rings but is broken into three pieces so it is not entirely clear if they all belong to the same individual); ♀︎ to 59. Body colour: Generally pale in colour but variable with some individuals darker brown in colour, adults with typical Cylindroiulus colouring though rather faint. Some darker dorsally. Juveniles with almost pink sheen. Overall body shape: Anterior constriction present but not strong. Head: Antennae pale in colour. Eye rows rather mixed up and difficult to read the lines, some individuals easier than others and some showing substantial variation between the two sides of the same individual. Several appear to have extra ommatidia. Stadia: ♂︎ VII-VIII; ♀︎ to IX. Body rings: Metazonites with strong and even striae, vaulting almost absent, posterior attenuation only slight, of apodous rings. Limbus not crenulate. Legs: Pale in colour, with accessory claw substantially longer than claw (Fig. 55). Telson: Preanal ring clearly projecting but not long and not extending beyond the anal valves (distinct from that of franzi which is very short), chestnut brown in some specimens (Fig. 47). With 2-3 pairs of setae, 1-2 ventrally and one on the telson projection but not at the tip and set back a short distance. Subanal scale not projecting, with 1 pair of setae. Anal valves with 1-3 pair of setae.

DIsTrIbuTIon
The additional records do not add greatly to the known distribution for C. londinensis (Fig. 60) which appears to be on the edge of its range on the Iberian Peninsula. noTes In a few locations only females were found but the shape of the telson coupled with the size of the specimens allowed placement within this species with reasonable certainty.
One Cylindroiulus londinensis had an expanded cheek plate and hook shaped first leg pair but one pair of normal walking legs on body ring 7. The combination of these characteristics is not suggestive of an intercalary male (these usually have a very short first pair of legs that are not hook shaped and a cheek plate intermediary between a mature male and a juvenile). Presumably this individual is just an aberrant example or a gynandromorph. It was not examined for vulvae as is now in the collection of S. Gregory. This female was exceptionally large in size with a body height of 5 mm. Probably 12 rows of ommatidia (= stadia XIII) and 45+1 body rings. The preanal process was distinct but not especially long. Another exceptionally large female was found at location (7) Leitza, area Kornieta. At this locality mature males of C. londinensis were also found, adding to the likelihood of these specimens being this species but just exceptionally large sized individuals.

Cylindroiulus cf. londinensis
For further discussion of this species see under Cylindroiulus sanctimichaelis below. (Leach, 1815) Material studied (2)  There are three samples where no males were found but the pale head, along with the overall size and telson/subanal scale shape indicated this was the correct species.

DIagnosIs
Small to medium sized species, with strong and down pointing telson projection and projecting anal valve, although not as pincer-like as in C. elosegiorum sp. nov. Generally well pigmented but with obvious pale head, first few rings and anal valves.

DescrIPTIon
Length: ♂︎ 16-18 mm; ♀︎ up to 24 mm. Body height: ♂︎ 1.2-1.5 mm; ♀︎ up to 2 mm. Body rings: ♂︎ 36-40 podous rings; ♀︎ up to 42. Body colour: Generally glaucous and dull, dorsal generally darker than ventral. Head and first 1, 2 or 3 rings usually noticeably paler (Fig. 62), becoming yellow in preserved specimens. Despite the overall pale colour there is often a darker band covering and linking the eyes (Fig. 63). Telson, mostly dark, with pale posterior margin and dorsal projection gradually fading in colour to straw/yellow (Figs. 62, 65). Colour in the juveniles sometimes appearing less contrasting. Overall body shape: Anterior constriction behind head but no posterior attenuation. Head: Antennae appearing relatively long and slender. Eye rows generally clear but final row often with reduced number of ommatidia and rather jumbled (Fig. 63). Stadia: ♂︎ 5RO = VI to 8RO = IX; ♀︎ up to 9RO = X. Body rings: Metazonites rugged and almost fluted, with close and strong striae sometimes giving a banded appearance, at least anteriorly (Figs. 64). Limbus not crenulate under light microscope. Legs: Legs with accessory claw surpassing claw (Fig. 66). Telson: Preanal ring with long and strong projection, strongly bent ventrally towards the tip. One pair of setae on the tip of the projection, another pair ventrally (Fig. 65). In smaller juveniles of stadium VI the telson projection appears relatively shorter and the down turn less pronounced (Fig. 67). Subanal scale projecting beyond the anal valves a relatively long distance but slightly ventrally, so less 'pincer-like' in appearance than C. elosegiorum sp. nov., with 1 pair of setae towards the tip. In smaller juveniles the projection is still present but relatively less well developed. Anal valves with 3-4 pairs of setae (Fig. 65).
Mature male secondary sexual characters: Mandibular stipites of mature males distinctly expanded and roughly square shaped. First pair of legs Although only females were found in Leitza the shape of the telson projection of this species is characteristic and it seems highly likely that the specimens found were this species.

DIsTrIbuTIon
This widespread north-western European species just extends into the Iberian Peninsula south of the Pyrenees. The current collection confirms its presence in this area and adds slightly to previous records (Fig. 61), but it does not appear to be common. Flagellum of normal length. Mesomerite short and squat, slightly shorter than C. sagittarius, rounded in posterior view. Opisthomerite chunky but posterior margin very membranous and therefore hard to see when glanced at quickly, so the opisthomerite appears at first sight more slender and simple in outline, thickened distally. (Compare Fig. 70 where the membranous area appears faint with Figs. 68-69 where it is drawn in). Membranous area on posterior margin of opisthomerite runs adjacent to the paracoxal process so they appear part of the same structure. Paracoxal process more obvious in posterior view and triangular in shape. Female first leg pair thickened and second also perhaps slightly enlarged.

relaTIonshIPs wITh oTher sPecIes
One of a relatively small number of Cylindroiulus species with strongly projecting sub anal scale as well as a strongly down curved and pointed telson. Others with this feature are C. apenninorum, C. elosegiorum sp. nov. (see above), C. gestri and C. propinquus all of which have different shaped gonopods and, as mentioned above under C. elosegiorum sp. nov., this does not appear to have any evolutionary implications.

ecology
The species was found in the leaf litter but, while all locations had trees present, many were described as being more open and/or a wood pasture and almost all were associated with a stream or, in one case, wet wood pasture.

lIfe hIsTory
The stadial spectrum (Fig. 73) indicates that males are semelparous, being mature just in two stadia (VIII and IX). The pattern for the females indicates a similar life cycle (Fig. 74).

DIsTrIbuTIon
Kime & Enghoff (2017) shows the range for this species along the whole of the Pyrenees and extending a little away from the mountains in southeast France and west in Spain along the coast. The current collection falls within the known range and added nothing new to the atlas maps. Recent experience suggests that it is fairly widespread, from Pau area in France, across the Basque country at least as far as Gorbeia in Alava although it does not appear to be particularly common. C. pyrenaicus has recently been found in the UK where it is likely to have been introduced, and also in several sites in northwest France (Gregory et al., 2018).
The current collection of C. pyrenaicus was compared carefully with the original description made by Brölemann (1897) and the specimens from ventrally, not extensively expanded (Fig. 77). First pair of legs in mature males small hooks. Keel of body ring 7 in mature males opening very wide.
Gonopods : Promerite: Medium length and fairly wide; no 'window'. Mesomerite: Considerably shorter than promerite, apex rounded. Opisthomerite: Very long and thin, expanded at proximal part to a small but distinct brachite and a membranous area between brachite tip and solenomerite. Smaller phylacum. Flagellum usually seen poking out of the tip of the opisthomerite. No obvious paracoxal process, just a slight broadening at posterior and base of opisthomerite.

relaTIonshIPs wITh oTher sPecIes
Gonopod shape is in this species is relatively unusual and it does not easily fall into a group with any other species. It has some resemblance to C. punctatus (in habitat as well as appearance) although lacks the long paracoxal process of that species.

assocIaTeD fungI
The species is notable for the fungus Rickia laboulbenioides found on the legs of some individuals (identified by Henrik Enghoff). In the current collection fungi were seen in 7 out of 21 mature males examined but only on two mature females. The fungus was found on the anterior legs, extending just beyond the gonopods. Some individuals appeared to have just a small quantity of fungi, but others had high densities making the legs appear 'furry '. ecology This species always seems to be found in and around decaying wood, fulfilling a role that in the UK is occupied by Cylindroiulus punctatus, although it perhaps has a preference for wood in the later stages of decay. Many of the sites where it has been found are wood pasture with old trees, especially beech (Fagus sylvatica L.). Cylindroiulus punctatus is also found in the Pyrenees/Basque area but seems less common.

lIfe hIsTory
Collections were made in two different times of the year (April and September). Mature males were only found in stadia VIII and IX and the breakdown in number was almost identical for each season (9 in stadia VIII and 3 or 4 in stadia IX) so the two collections were combined to produce the stadial spectrum which is shown in Figs. 82-83. This shows that the males are clearly semelparous, in the females however, while there is a peak in stadia VIII and IX, a small number were found in stadia X and XI hinting that either some females mature later or that a small number may moult following reproduction and therefore may have the capacity to be iteroparous. Navarra: Leitza, Ariz Mendiak, at area "Kornieta", 43.0698, -1.8823, 910 m, 4.2.2010

DIagnosIs
A medium sized species, chestnut brown, usually darker dorsally and often with paler head, but with dark band across the ommatidia patches. Large number of body rings. Strong, prominent and straight telson projection. Opisthomerite of gonopods unique in shape and very elongated with the flagellum projecting from the end.

DescrIPTIon
Length: ♂︎ 18-22 mm; ♀︎ up to 19 mm. Body height: ♂︎ 1.0-1.4 mm; ♀︎ up to 1.9 mm. Body rings: ♂︎ 44-54 podous rings; ♀︎ up to 57. Notable for the large number of body rings and the large number of apodous rings, even in mature individuals. Mature males had between 2 and 6 apodous rings. Body colour: Mid/chestnut brown, generally paler ventrally, with darker mid dorsal pattern (Fig. 75). Head often paler but with darker band extending across the ocular patches and the 'forehead'. Sometimes collum and anal valves paler too. Overall body shape: Anterior constriction not obvious in dorsal view, although strangely narrowed ventrodorsally. Body more or less parallel sided but apodous rings (usually high in number) reduced in height relative to rest of body. Head: Antennae generally pale with darker marks. Eye rows mostly clearly readable, ommatidia barely reduced in number. Stadia: ♂︎ VIII or IX; ♀︎ up to XI. Body rings: Metazonites with striae strong and regularly placed, quite wide apart. Limbus not appearing crenulate. Legs: Generally paler in colour. Accessory claw extending beyond claw to roughly 25% of the claw length. Telson: Preanal ring with very strong, level and pointed projection (Fig. 76). One pair of setae at end of projection and one pair ventrally. Subanal scale not enlarged and bearing 1 pair of setae. Anal valves with 3 pairs of setae.
Mature male secondary sexual characters: Mandibular stipites of mature males smoothly expanded and curved

DIsTrIbuTIon
Cylindroiulus sagittarius was confined to the Pyrenees and north-west Spain (Kime & Enghoff, 2017) and the current collection does not extend the range in this area. It has recently been found in the UK (Gregory & Owen, 2019) where presumably it has been introduced.

DIagnosIs
A chunky, dark brown species, similar to C. caeruleocinctus but with differences in the shape of the opisthomerite of the gonopods. Larger specimens have a distinct 'golden glow'.

DescrIPTIon
Length: ♂︎ 21-23.5 mm; ♀︎ up to 39 mm. Body height: ♂︎ 1.4-1.9 mm; ♀︎ up to 2.9 mm. Rings: ♂︎ 39-44 podous rings +2 apodous; ♀︎ up to 50+1 rings. Body colour: Chestnut brown metazonites, alternating with darker brown prozonites, slightly paler ventrally, some markedly so. Many specimens with a gold sheen to the metazonites, not as strongly metallic as C. caeruleocinctus but larger specimens with a 'golden glow' Fig. 84, collum generally dark in preserved specimens. Legs yellow brown. Overall body shape: Anterior constriction present, with distinct narrower 'neck'. Posterior attenuation quite strong, posterior quarter to fifth of body length noticeably narrower. Head: Antennae brown. Eye rows apparently clear and easy to 'read' but, in reality rows rather muddled so difficult to be sure of exact number in most mature specimens. Stadia: ♂︎ probably mostly 9 rows = X but up to 11 rows = XII; ♀︎ up to 11 or 12 rows = XII-XIII. Body rings: Metazonites with striae fairly strong, even and relatively widely spaced. Limbus not crenulate. Legs: Brown, accessory claw slightly longer than claw. Telson: No projection beyond anal valves. One pair of setae dorsally and one pair ventrally (Fig. 85). Subanal scale normal and not projecting. With one pair of setae. Anal valves usually with three pairs of setae but ranges from 1 to 5 pairs.
Mature male secondary sexual characters: Mandibular stipites of mature males relatively strongly projecting both anteriorly and ventrally. First pair of legs in mature male forming a small tight hook. Body ring 7 in mature males expanded ventrally to a distinct keel with anterior gonopods of some males projecting through the gap in ring 7.
Gonopods : Promerite: Long and slender. Mesomerite: Slender, shorter than promerite, blunt tipped. Anterior gonopods appearing identical to caeruleocinctus. Opisthomerite large, angular brachite, slightly convex in outline (C. caeruleocinctus is generally slightly concave). Opisthomerite more prominent and irregular in comparison to C. caeruleocinctus. There is an obvious division between the solenomerite and the phyllacum (as in C. caeruleocinctus and different to C. londinenis   difficult to determine than expected, especially for older individuals. The graphs include data for some individuals in which a 'best guess' has been made. The stadial spectrum for the males (Fig. 85) shows a relatively tight group of mature stadia and that of the number of podous rings (Fig. 86) seems to substantiate this. The females show a wider range of both stadia (Fig. 87) and podous rings (Fig. 88), not only through the inclusion of juveniles but also to the right of the graphs. That for the stadia is bimodal and this indicates that while it seems likely that the majority of males reach maturity and die following reproduction, some of the females may well survive to reproduce more than once.

DIsTrIbuTIon
This species has a widespread but scattered distribution within Spain (Fig. 91), from the foothills of the Pyrenees to Andalucía and down the eastern side of the Iberian Peninsula. There seem more records in the north, but this may be the result of greater collecting effort in these regions. Despite it being found frequently in the current collections the known range has barely been increased. The specimens from these collections were very similar to each other and clearly fall in the C. londinensis group (see below), probably closest to C. sanctimichaelis. In terms of size and ring number they fit within C. sanctimichaelis range (Figs. [16][17], although some were towards the low end of the height range. The eye rows appeared rather more regular, and the males appeared to have slightly longer legs in relation to the body size (Fig. 89) although this was clearly an illusion. Body height divided by total mid-body leg length (using the method described in Enghoff (1982) gave figures of 1.38 and 1.35 for males from Zaragoza and Guadalajara respectively and 1.33 and 1.40 for males from Parque Santa Lucía and Oteo respectively. The brachite, solenomerite and phylacum of the gonopods are all in a line (Fig. 90) and the brachite and phylacum are both large, thus resembling sanctimichaelis. However, the phylacum is more pointed distally and more rounded in overall shape, with less distinction between it and the solenomerite. The margin of the brachite also has a flatter profile. It is not easy to determine if these specimens are sufficiently consistently different to consider them a separate species or if they are individuals within an overall variable species. Genetic studies would be useful to help elucidate. (Blower 1985)). High paracoxal rim. No paracoxal process, although perhaps a small 'nick' in the posterior line of the gonopods basally. ecology Vicente (1981) noted that this species was very abundant in Quercus ilex forest, amongst the leaves. For Serra et al. (1997) it was the second most abundant Julid species in a sclerophyllous forest consisting of Q. ilex with an understorey of Arbutus unedo. Serra et al. (loc. cit.) recorded a mean annual density of 35.22 individuals per meter squared for this species with peaks in later summer and autumn. He also recorded a preference for the intermediate and deeper soil horizons and that it was notable for the high proportion of females relative to males. Sampling by Serra et al. (1997) was carried out by taking soil cores and extraction using Berlese-Tüllgren devices. In the present collection it was found in the leaf litter of various woodlands, but the most extensive collection was made in a Quercus ilex forest where it was very abundant.

lIfe hIsTory
The current collection had a good number of specimens, so opportunity was taken to count eye rows and podous segments in order to attempt to shed a light on the life history of this species. Unfortunately, the eye patches were composed of a large number of ommatidia and, while appearing from a distance to be regular and easy to count, they were actually more   Kime & Enghoff (2017); el rojo, muestra la nueva localidad del material actual.

C. londinensis. All have a relatively large body size for
Cylindroiulus and similar gonopods with a large and obvious brachite, a phyllacum of varying size, a small or absent paracoxal process but a high paracoxal rim.
In the present collection, several people (including myself) initially misidentified C. sanctimichaelis as relaTIonshIPs beTween C. sanCtimiChaelis anD sImIlar sPecIes Cylindroiulus sanctimichaelis is one of a group of seven very similar species all found in the Iberian peninsular: C. sanctimichaelis, C. caeruleocinctus, C. caramelos sp. nov., C. chalandei, C. finitimus, C. ibericus, 1913 and between them, with C. sanctimichaelis seeming to have been quite common and C. finitimus rather rarer in her region. Vicente's (1981) figures of gonopods show the brachite of C. sanctimichaelis much longer than the solenomerite and phyllacum, (like Mauriès's drawings of C. londinensis) and a phyllacum more pronounced, both in size and definition from the solenomerite, than in C. chalandei and C. finitimus. The brachite was also flat topped and not pointed in C. finitimus, a feature shown in some of the illustrations in Mauriès (1964). Vicente (1981) also lists C. finitimus as being slightly larger than C. sanctimichaelis but with fewer body rings. Vicente & Ascaso (1990) subsequently made a detailed study of the ecology of C. ibericus suggesting that there were no difficulties distinguishing it as a species. Table 1 shows the characteristics of these similar species with details taken from the literature, both the original descriptions and previously published comparisons. Some information was added from personal observation where specimens were available, for example the shape of the tip of the mesomerite in posterior view, something indicated in the literature as being potentially of help in distinguishing some species. The relative lengths of the three main structures of the opisthomerite (brachite, solenomerite and phylacum) have been used previously by authors to distinguish the species and this does seem helpful for some of them. See also Figs. 92-104 which illustrate the gonopods of the various species. Specimens of C. finitimus in NHMD were checked by Henrik Enghoff and a comparison made with photographs from the current collection. They were found to clearly differ and had a much smaller phylacum. No specimens of C. chalandei or C. ibericus were available for study.
In summary: • C. chalandei has a pointed telson projection and the overall gonopod shape seems reasonably distinct (no specimens seen) (Fig. 98) • C. ibericus also seems to have a distinct gonopod shape with a large and angular brachite and a very slight phyllacum (Figs. 101-104). The shape of the brachite depends at least partly on orientation but more detailed sketches of the gonopods by Bröleman, do show the brachite to be noticeably more angular than any other specimens seen. The mesomerite also looks a little strange in some illustrations, for example that from Attems (1927) (Fig. 104). No specimens were seen as part of the current study, however. • C. londinensis usually has a distinct telson projection (see also , although in C. finitimus it appears also to be variable and could perhaps be similar in some specimens; it is distinctly stout for the body length. • C. finitimus appears to have a much reduced phylacum in comparison to others in the group C. caeruleocinctus and it was only when looking more critically at the gonopods that it became evident that they were closer to C. sanctimichaelis. The specimens from the Basque region appear to be slightly smaller, and generally with slightly fewer podous rings, than previous descriptions. In addition, the number of pairs of setae on the anal valves was more variable and ranged from 1 to 5 (with a mean of 3.2). However, careful examination of the gonopods revealed them to be most like C. sanctimichaelis. Cylindroiulus sanctimichaelis was described by Attems (1927) who distinguished it from the similar C. ibericus, a species described previously by Brölemann (1913). Several previous publications have also compared groups of 2-3 species such as Mauriès (1964) for C. finitimus, C. londinensis andC. caeruleocinctus andVicente (1981) for C. chlandei, C. finitimus and C. sanctimichaelis. Older literature can be misleading because of the confusion with C. teutonicus (now regarded as C. caeruleocinctus) and that of C. londinensis/C. finitimus (C. finitimus was previous considered a subspecies of C. londinensis). Attems (1927) compared C. sanctimichaelis, C. ibericus and C. teutonicus (i.e. C. caeruleocinctus) and considered them to be very similar. He noted that C. sanctimichaelis had a clear notch in the gonopod between the brachite and the solenomerite. In his key he separates C. caeruleocinctus from the other two on the basis of the presence of the teeth on the apex of the solenomerite and then C. sanctimichaelis as having 3 pairs of setae on anal valves (C. ibericus having 5-6 pairs), and shape of the posterior gonopods. However, in the current collection the number of setae on the anal valves of what is here considered to be C. sanctimichaelis is variable and there are often more than 3 pairs, however the gonopods do look more like those figured by Attems as C. sanctimichaelis. While the number of pairs of setae on the anal valves may be indicative of particular species it may not be a reliable character and sometimes the setae are worn off and the point of insertion difficult to see. Mauriès (1964) looked in detail at C. caeruleocinctus, C. londinensis and C. finitimus and decided they were all good species, despite Ribaut (1905) originally describing C. finitimus as a variety of C. londinensis. Mauriès (loc. cit.) separated them by the telson shape (in C. londinensis) and the relative length of brachite to the solenomerite (and phylacum). He also noted the clear division or notch between the brachite and the solenomerite in C. caeruleocinctus and C. finitimus. A character also used by Blower (1985) to separate C. caerueocinctus from C. londinensis. Mauriès (1964) concluded that C. finitimus is a distinct species with constant characters, a geographic division (higher altitudes and in the valleys of the central and eastern Pyrenees) and a well-defined habitat. Vicente (1981)  chalandei (Ribaut, 1904) finitimus (Ribaut, 1905) ibericus Brölemann, 1913 londinensis (Leach, 1815) sanctimichaelis Attems, 1927 Information from Blower (1985); Mauriès (1964) Current description Ribaut (1904) Ribaut (1905); Mauriès (1964); Vicente (1981) Brölemann (1913) Blower (1985); Mauriès (1964) Attems (1927)  Characters also examined were the shape of mesomerite, teeth on solenomerite, the anterior margin of brachite and the shape of paracoxal process. In general, the differences between species for these characters were not helpful in distinguishing them except where stated in the 'additional comments' line of the In conclusion, it seems likely that all the six previous described species, plus that newly described here, are reliable and can be distinguished, but they are challenging to determine. Carefully scrutiny is needed to separate them and there is some variation within some of the apparent species as here understood. It may be that there are fewer, more variable species, with C. caerueocinctus and C. sanctimichaelis (for example) two ends of a spectrum, however they are here regarded as two separate species. The specimens from Zaragoza and Guadalajara seem a little different from the 'typical' C. sanctimichaelis as seen in other specimens in this collection, but not different enough to be confident in ascribing them to a new species. This approach was supported by the graphs showing rings vs body height, where they fall within the current variability of C. sanctimichaelis .
It is clear that many previous workers have struggled to determine species within this group, and it would benefit from combined genetic and morphological study to explore the species concept within this group and to look at the variation within some of the more widespread species, both within the Iberian Peninsula and across other parts of northern Europe.
In light of the number of Iberian species of millipede that have been found in Britain in recent years (see under C. pyrenaicus and C. sagittarius above) it is potentially possible C. sanctimichaelis, which seems quite a widespread species in Iberia, may also be found, so specimens of C. caeruleocinctus should be scrutinised carefully. Attems, 1952 Figs. 105-112

DIagnosIs
A small species, lacking any telson projection or distinguishing external characteristics, but with posterior gonopods unlike any other species in the genus so far described.  Mauriès (1964) (when discussing the first three species) it is not always easy to determine. • C. caramelos sp. nov. has a large number of body rings for the body height and anterior gonopods relatively large in comparison the posterior ones. • C. londinensis and C. sanctimichaelis are illustrated by Vicente (1981), Mauriès (1964) and Blower (1985) as having a brachite substantially longer than the solenomerite, (as opposed to C. finitimus and C. caeruleocinctus) where it is of similar length. This is less obvious in the specimens illustrated here and orientation is important for this feature.
Other characters that might be worth considering in future include the colour. Larger preserved specimens of C. sanctimichaelis observed here appeared to have a 'golden glow' which is rather different to the metallic copper colour seen in C. caeruleocinctus, but the literature suggests that other species have a tendency to be golden too (such as C. finitimus and C. chalandei). C. caramelos sp. nov. (see above) is apparently black with white legs. The ventral margin of the brachite could also be useful as it can be moreor-less flat, or with a convex or concave curve.
The distributions of the species may also help. Kime & Enghoff (2017) show that C. ibericus has a very narrow distribution in Huesca. Cylindroiulus finitimus seems also only to have been found at high altitude in the Pyrenees and Mauriès (1964) referred to it as an altitudinal variant of C. londinensis. The locations of both of these species fall within the range of C. sanctimichaelis which is considerably more extensive. Cylindroiulus londinensis and C. caeruleocinctus are much more widely distributed across Europe but it seems likely that they may have originated in Iberia. Cylindroiulus caramelos sp. nov. has only been found from area of mountains near Tarragona connected to the Sistema Iberia but not part of the Pyrenees, however also within the range of C. sanctimichaelis, but it may also be common further south around the Ebro delta (H. Reip pers. comm.) It is possible that some of these species, particularly C. caeruleocinctus and C. sanctimichaelis, form a complex in Iberia but that one morph has become more widespread in northern Europe and remained relatively consistent in form. This situation is perhaps similar to that of Ommatoiulus moreletii (Lucas, 1860) which seems to be variable in Spain and Portugal but those introduced to Australia are less so (Baker, 1984). Another Ommatoiulus species, O. diplurus (Attems, 1903), also seems to be polymorphic in Spain, with considerable variation in gonopod structure (Akkari & Enghoff, 2012).  Ribaut (1904), mesal view. 99. C. finitimus redrawn from Vicente (1981), mesal view. 100. C. finitimus redrawn from Mauriès (1964), mesal view. 101. C. ibericus redrawn from Brölemann unpublished, mesal view. 102. C. ibericus redrawn from Brölemann unpublished, lateral view. 103. C. ibericus redrawn from original description by Brölemann (1913), mesal view. 104. C. ibericus gonopod redrawn from Attems (1927), mesal view.

ecology & lIfe hIsTory
There are insufficient specimens from the current collection or in the original description to make any comments about the ecology or life cycle of this species.

DIsTrIbuTIon
The type specimen was found in El Pardo, Madrid Province, just north of the city. Guadalix de la Sierra is just a little further north. Kime & Enghoff (2017) show an additional location near Zaragoza collected in dry gypsum ground with planted Pinus halepensis Mill., probably not far from Monegros which is just east of Zaragoza, thus the two specimens add slightly to our knowledge of the distribution of this species (Fig. 112). Both referees commented that this species is locally common in many areas in and around Madrid Province.

Cylindroiulus indet.
Two collections of females and juveniles could not be identified to species and could be C. latestriatus or C. britannicus (Verhoeff, 1891 anteriorly. Photographs of a live animal provided by the collector (Fig. 105) show a pretty banding pattern at least dorsally. Overall body shape: With little variation, anterior constriction slight and no posterior attenuation. Head: Antennae brownish in colour. Eye rows making a full equilateral triangle and rows reasonably 'readable'. Stadia: ♂︎ X-XI; ♀︎ up to XI or XII. Body rings: Striae regular and vaulting negligible. Limbus not crenulate. Legs: golden with brown markings, darker anteriorly. Accessory claw of similar length to claw. Telson: Pre anal ring lacking dorsal projection, not dark in colour but with darker margins. Bearing one pair of setae ventrally and another pair on the tip of the telson (Fig. 106). Subanal scale bearing one pair of setae. Anal valves darker basally, becoming more orange/brown posteriorly. With three pairs of setae. Mature male secondary sexual characters: Mandibular stipites of mature males expanded ventrally, with almost vertical anterior margin, (Fig. 107). First pair of legs in mature male tight hooks with slight 'knee'. Keel of body ring 7 in mature male projecting ventrally to a low ridge around a wide ventral opening.
Gonopods : Promerite clearly longer than mesomerite, with no perforation or 'window'. Flagellum of normal length. Mesomerite flat topped, parallel-sided and fitting into a slight 'cavity' in the promerite. Large and high paracoxal process. Opisthomerite distally divided into two parts with large gulf between them. Solenomerite expanded along the length by a phylacum. Brachite large and bird-like in shape with a 'beak' pointing towards the solenomerite opening. The shape of the opisthomerite is characteristic but much more complex than that illustrated in the original description. No obvious paracoxal process but slight 'bump' on posterior margin of opisthomerite. relaTIonshIPs wITh oTher sPecIes Attems (1952) described C. unciger in the subgenus Alpicylindrus, together with C. festai Manfredi, 1937, C. limitaneus (Brölemann, 1905), C. tricuspis Verhoeff, 1932 andC. aostenus Verhoeff, 1932, which are all Italian and have gonopods that look nothing like the those illustrated by him for C. unciger. Attems (1952) also described differences between the gonopods of C. unciger and those of C. bouvieri which he considered a similar species. Again, the gonopods in the description of C. bouvieri by Brölemman (1896) (= C. parisorum) look nothing like those illustrated by Attems (1952) for C. unciger.
The present specimens are in accord with the original description in terms of size and telson shape although the type was black in colour and had 56 body rings. The illustrations of the gonopods in the original description are very simplistic but the overall shape, the relative length of the promerite to mesomerite and the high paracoxal process all accord with the current specimens. The location in the province of Madrid also seems to indicate that they are this species.   Kime & Enghoff (2017); los rojos, nuevas ubicaciones del material actual.