Additional type and other notable specimens of Mollusca from the Montagu Collection in the Royal Albert Memorial Museum & Art Gallery, Exeter

This paper completes the review of the Col. George Montagu collection of shells in the Royal Albert Memorial Museum & Art Gallery, Exeter. A further twenty-one lots of type material were discovered bringing the total number of type lots to ninety-four. A number of other taxa that are of historic and potential nomenclatural significance are discussed. Specimens that relate to taxa of authors other than Montagu were isolated and consideration was given to the many non-British taxa that Montagu included in his works. To complete the study a list of all species represented in this collection is given. Key Words


Introduction
In 2017 (Oliver et al. 2017) published a catalogue of the extant type material available for molluscan species described by George Montagu in 1803Montagu in , 1804Montagu in , 1813 and 1816. Following on from this further curatorial work was carried out in the Royal Albert Memorial Museum & Art Gallery, primarily to investigate specimens that were described by contemporaries of Montagu notably Richard Pulteney, Edward Donovan, William Turton, Thomas Rackett and E. Mendes da Costa. Of special interest here were the many non-British species included by Montagu. During this work a number of type specimens reported as missing by Oliver et al. in 2017 were located. Given that the Montagu collection had been used as a source of display material, at least twice, and remounted during the process, since its acquisition it is not surprising that the original data had become obscured. The first part of this paper reports on these newly discovered type specimens.
The remainder of the paper analyses some notable, but not type specimens and the non-British shells extant in the Montagu collection. Many species that Montagu included had been noted by previous British authors such as E. M. da Costa (1778), Pulteney, (1799) and Donovan (1804). Our aim was to compare Montagu's specimens with the published accounts to ascertain how accurately this material had been identified and could it be used as a proxy for type material for which little now remains.
Finally we will give an overview of the entire collection as it now exists in RAMM by presenting a list comparing the names used by Montagu and their current identifications. This will give further insight into how accurate Montagu's works were.   Montagu, 1803 (EXEMS Moll4149) and Helix paludosa da Costa, 1778 (EXEMS Moll4148), no shells attached. 6.2 3 shells identified as Vallonia excentrica Sterki, 1893, one can be associated with paludosa card (arrowed). 6.3 4 shells tentatively identified as Vallonia costata (Müller, 1774). 7. Bulla diaphana Montagu, 1803 [Trivia arctica (Pulteney, 1799)]. 7.1 "Rowley" box for lot 4311 with 2 shells, Syntypes. 7.2 Montagu hexagonal blue card mount with 1 shell attached. 7.3 Apertural and abapertural views of largest but damaged syntype, EXEMS Moll4311a. 8. Helix lacuna Montagu, 1803[Lacuna parva (da Costa, 1778]. 8.1 Montagu black hexagonal card mount with 5 shells, bottom right (arrowed) is Lacuna vincta. 8.2 basal, apertural and abapertural view of 1 of 4 syntypes, EXEMS Moll4189. 9. Nautilus lacustris Montagu, 1803[Segmentina nitida (Müller, 1774]. 9.1 "Rowley" box with blue card and 4 shells, Syntypes, EXEMS Moll4106. 9.2 Montagu blue hexagonal card mount and printed label. 9.3 apical, basal and lateral views of syntype EXEMS Moll4106a. 10. Helix lutea Montagu, 1803[Radix balthica (Linnaeus, 1758]. 10.1 Old box and original hand-written label for lot 4117. 10.2 Apertural and abapertural views of unique syntype EXEMS Moll4117. Montagu G 1803. Test. Brit. Part 2. p. 428-429, Tab. 13 Pulteney (1799) and restated by Maton & Rackett (1807) and then by Montagu, all referrring this taxon to Linnaeus.  was reluctant to accept this species as British but it would appear that specimens were in circulation in the early 19 th century. The Lyons collection, (dating from around 1810) in Tenby contains two lots both with Weymouth (from Miss Pocock) as their origin, and in keeping with the original record of Pulteney. The RAMM label gives the source as "Mr Dillwin, Bantry Bay" and this would coincide with Dillwyn's visit to Ellen Hutchins the Irish botanist in 1809. The RAMM and Tenby shells are in good condition but it is not known if they were collected from sites in the UK or from collections brought back from its native Caribbean. Jeffreys (1864, p. 299)  Montagu stated that da Costa must have been mistaken in regarding it as common on west coasts and goes on to say that neither he nor Pulteney ever found it. The provenance of the two shells in RAMM is not known, the species is widespread in the Caribbean pintado Turbo Figure 47. Littoraria pintado (Wood, 1828) [Littorinoidea, Littornidae]. EXEMS Moll4307, 2sh. This species is not included in any of Montagu's publications and it is puzzling, as it was not described until 1828 by Wood. The mount and label are typical of all of the Montagu mounts in Exeter. This is the only original label in the Montagu collection that bears a species name erected after the death of Montagu. There are a number of unpublished names on similar labels and one can only surmise that this is a manuscript name adopted at a later date by Wood. These shells have the operculum in place suggesting they were collected alive, which is unlike the majority of the exotic shells in the collection as most are worn and have some degree of incrustation upon them. It is noteworthy that the specimens of Echinolittorina ziczac (Gmelin, 1791) in the Montagu collections also have the operculum in place and appear live collected.

List of taxa present in the Montagu collection in RAMM Exeter
The following is a summary of the entire Montagu collection in RAMM. The collection contains few duplicate lots and was perhaps a voucher series retained by Montagu's wife Eliza, following the sale of the major part of the collection to the then British Museum. All lots are listed including duplicates. The first line in bold text is the name as used by Montagu in his publications and in these he gave the authority but not the date, the second in normal text is our current identification using nomenclature from MolluscaBase. A number of specimens bear label names that are not present in any of Montagu's publications or only bear labels that are more recent. Type material is denoted by the prefix superscript T. As noted in our earlier paper recognising the provenance of lots in the Jeffreys collection is difficult, as the original labels have been removed. The relevant portion of this letter is reproduced here with a transcript. It seems clear now that the missing lots will not be found in RAMM. "Some years ago, when I was in Exeter, I took the trouble to go thoroughly into the matter of the Montagu spp..-on what Jeffreys reported in J. of Conch. II, 1-4. Unluckily the shells were all sent to J. with the help of the Curator. I ran down the entry in their records of the sending of the parcel to Jeffreys (I think it was in 1878). But there is no record of their return: the old scoundrel pinched them + no doubt they are now in Washington. I looked carefully through their collection and none of the spp mentioned in J. C. are forthcoming. I think the matter ought to be taken up and restitution demanded from the U. S. Nat. Mus. If properly and courteously done I think the things would be returned"

Tellina flexuosa Montagu
The remaining collection is contained in 316 lots covering register numbers 3639-4134 but no longer representing the number of shells noted in the register and it would appear that many, especially of the smaller shells, have becomed dislodged from the mounts and lost, some time prior to 1979. The majority of the shells verified by Brind in 1979 are present. The number of type bearing lots is now 92.
Although Montagu continued to publish up to his untimely death our review suggests that he was still studying shells and recognizing species not included in his earlier works, such as Otina and Roxania.
We have examined 23 non-native species included by Montagu (1803, 1808) but not described by him. Most the locality. Nowhere is there any mention of this being a terrestrial species, typically found in hot houses. Once introduced into the British literature it would appear that such species were in demand by collectors and shells were then sourced but spuriously given the collection site as that of the original finding.
It must also be reminded that wooden shipping of the 18 th century was a much different environment for possible translocation of organisms. The ships built up a much larger volume of fouling organisms enabling non-attached species such as snails to remain protected. This could explain how Montagu had two species of Caribbean littorinids with intact opercula in his collection.
The vast majority of Montagu's determinations are accurate according to the literature of the time. Montagu is honest in his admissions of being uncertain about certain taxa e.g. his confusion with species of Thracia. Although the nomenclature is often mistaken and Montagu was prone to introducing new names for known species he seldom splits species on minute differences. The RAMM collection is however far from complete and lacks may of the minute marine and terrestrial species such as those in the families Pyramidellidae and Vertiginidae. In some cases the extant material does not conform with the description given such as Turbo sexdentatus where none of the shells have six teeth and do not conform with the current concept of being a junior synonm of Vertigo antivertigo. This indicates that caution must be used when considering the type status of specimens and questions how consistent Montagu was in his identifications especially of these smaller species.
were reputedly collected from the south coast of England, mainly Dorset, and many by Mr Bryer although some were given by Pulteney, Boys and Laskey.These shells for the major part have been identied correctly suggesting that there was consistency between these early collectors. The underlying issue with the non-British shells included by Montagu is not their identity their provenance as discussed by Oliver et al. (2017). Although the source as ballast was considered by Oliver et al. (2017) it was the co-occurrence of species found on the south coast of England and in Scotland that gave rise to the suspicion that the shells were not found naturally. The species are exclusively from the Caribbean region and all but two represented by empty shells showing signs of being long dead. If there had been an underhand attempt to introduce exotic species then one might expect such shells to be from numerous localities and to be in good condition. The Caribbean (West Indian) origin fits well with the bulk of the trade during the late 18 th century and given the number of ships lost at this time (Boult 2003) it may not be unreasonable to assume that access to the ballast of a wrecked ship could produce this assemblage of shells. Where suspicion still arises is the co-occurrence of these species reported from Scotland by Laskey (1811) and in  list of North British Shells.
The lack of accuracy of records kept by collectors may also have added to our suspicions. For example Pulteney (1799) records Subulina octona from Weymouth but gives no details as to where. Weymouth is repeated by Miss Pocock for the shells sent to Lyons in Tenby and the shells collected by Dillwyn give only Bantry Bay as