Article © Julian Dignall, uploaded March 12, 2022.

In the beginning, there were not many Corydoras. Of course, Corydoras have been around in the wild for tens of millions of years, but they have only really been documented by fishkeeping humans in the past century. In other words, despite being first collected in the late 18th century, when looking before the 1960s or so, only a handful of them appeared in books usually with old names (e.g. Corydoras microps, C. myersi or Corydoras funnelli) being used for common species we know today. And while many incurred name changes or suffered from misidentifications that are clear today, from the very beginning, Corydoras arcuatus was a species aquarists and scientists alike agreed on.

It wasn't hard after all: while some corys have experts cross-eyed with the presence or absence of a spot here, or a pigmented fin ray there, the arched cory was simple. Even into the 1980s when imports of the larger, long snouted Corydoras narcissus showed up, this was clearly a bigger, boisterous big brother and there was not a lot of confusion. Everyone knew what Corydoras arcuatus was and while C numbers emerged and evolved into CW numbers as decades passed with Corys ever more legion, Corydoras arcuatus remained a banker. The arched cory (or the superb German name, Stromlinienpanzerwelse basically meaning streamlined mini tank (armoured) catfish), the fish we all knew and saw in the hobby matched the description, and the picture therein, to a tee.

This is where Steven Grant comes in, around the time I was starting PlanetCatfish.com in the late 1990s, Steve was describing Corydoras kanei and C. crimmeni in Aquarist and Pondkeeper Magazine. The former named for his son, and the latter for Mr Oliver Crimmen, the Curator of the fish collection at the Natural History Museum (BMNH) in London. A few years later, and pretty much every Catfish Study Group convention and open show, rain or shine, Steve was there - one of the catfish hardcore. The faithful attendees learned many things at those excellent events, one standout lecture from Markous Alexandrou was the introduction of lineages into the mega genus, Corydoras. It got me thinking about something I'll discuss another day, but what it also did was formalise what aquarists had seen in their corys. That while some shared a colouration pattern, their body shapes and behaviours were different. What Steve did was take this one, big, step further.

There is an CAGB article that Steve wrote in 1993 where he queried what the C. arcuatus was: a round nosed species, or a long nosed one. The BMNH began to put images of its collections online early in the new millennium, with the images of Corydoras arcuatus being available to all in 2003. The description of another arched cory, Corydoras urucu in 2009 overlooked the issue despite examining the images. However, having already described two Corydoras with a similar colour pattern to each other but with different body shapes, when he saw this image described as the holotype of Corydoras arcuatus, Steve identified this was not a round nosed specimen and so began a process culminating in the description of C. granti in 2020.

Steve researched Corydoras arcuatus and published, in 2014, the key article in the Catfish Study Group's journal identifying what all others had missed for 75 years. He hypothesised that the type photographed by the BMNH was a different species from the Corydoras arcuatus everyone thought they knew. One thing I've found over the years is it's easy to say a particular specimen is not one particular species or another, but it is often not easy to say what that particular misidentified specimen actually is. Steve tackled this by discussing all the arched corys which include C. narcissus Nijssen & Isbrücker, 1980 (Rio Purus system, Brazil), C020 (Rondônia, Brazil), CW036 the "Super Arcuatus" from the Rio Madeira, Brazil and Corydoras urucu (Rio Urucu basin, Rio Solimões system, Brazil) and also, C019 (Brazil), C098 (Brazil), C100 (Rio Negro, Brazil), C. evelynae Rössel, 1963 (Upper Rio Solimões, Amazonas, Brazil), CW006 (Peru). Steven went on to propose that the C. arcuatus was the same as CW036. While listing out all the arched corys, one must also mention CW006 was also recently described as Corydoras bethanae, Bentley, Grant & Tencatt, 2021 which was recently named to honour Steve's daughter.

Five years on from the CSG Journal article, Tencatt, Lima, & Britto published, "Deconstructing an octogenarian misconception reveals the true Corydoras arcuatus Elwin 1938 (Siluriformes: Callichthyidae) and a new Corydoras species from the Amazon basin" in the Journal of Fish Biology. In my mind this is one of the most thorough and collaborative Corydoras descriptions I've read (and there are a few!) C020 was named as C. granti recognising Steve's work. I really enjoyed the paper as it covered a lot of natural history, behaviour and the reader could see the collaboration between all those involved leaping off the page.

Some questions still gnawed away at the edges of all this in my mind. Why was C020 created when we "all knew" what C. arcuatus was? It turns out that the specimens were juveniles and their pattern had not fully formed. Why was what we now call Corydoras granti photographed in Elwin's paper but Corydoras arcuatus deposited in the BMNH? What was the other record of this species there? Why was the description written half a year or more before the "first" specimen was deposited at the BMNH? What happened to the other specimen mentioned in Elwin's paper. But the big one really, was, how did this happen and, by inference, could history repeat itself?

Fishkeeping is a large group of hobbies including specialisms not just in terms of groups of fishes kept but interests as wide apart as photography, chemistry, showing fishes, fishroom construction, breeding live foods - all sorts of things. One of my side interests is the history of it all. Facts like Neon Tetras were amongst the passengers on the ill-fated Hindenburg Zeppelin on its last successful voyage before the disastrous trip from Frankfurt to New Jersey in 1937 send me off to sitting with books for hours to glean an understanding of what went on and consider what I read through the fishkeeping optic. And so, I went through older literature in search of arched Corys.

Steve had already done a good job of researching which scientists had examined these species and all I wanted to do was check if further images existed in those works, possibly of the holotype in the BMNH. I hoped to find a picture in the scientific literature from 1940 (Gosline) onwards but was unsuccessful after reviewing the works of Nijssen & Isbrücker, Castro, and Britto et al. It's clear the holotype was examined, but even Nijssen & Isbrücker's seminal works while featuring detailed morphometrics, which are very useful when comparing species side by side in tabular or visual form, they are nigh on impossible to otherwise visualise in terms of the whole fish. So, no clues found there. The hobby literature, especially post 1970, is littered with plates and then photographs but again what they show is almost always Corydoras granti and it is clear they worked from the photo in the description.

The Innes book, or "aquarium bible" of the time, Exotic Aquarium Fishes was around well before, during and after the introduction of the arched cory. It was the first English language book to depict the arched cory. It has some interesting passages noting it was collected by Rabaut and was initially called the Tabatinga Cory because of its origin - a popular Corydoras collection locale even to the present day. The original series published by Innes (as opposed to the cheap reproduction published by Axelrod/TFH after Innes failed to renew the copyright) has high quality plate which shows, to my mind, Corydoras urucu because of the round nose and incomplete arched stripe. The name Tabatinga catfish continues in use (Exotic Tropical Fishes, TFH into the early 80s). From then on, the major fish books of the 50's onwards show what is now Corydoras granti.

A notable exception is one of the first colour pictures of the arched cory which appears in Sands' Catfishes of the World and likely shows Corydoras arcuatus as we now know it. It is not known if that author visited the BMNH and examined the holotype, but it is possible given the significant research put into the Catfishes of the World series of books. On balance however, this correct identification is likely accidental as the author also reviews the prior published large set of Corydoras pictures in Emmens & Axelrod's, "Catfishes" and notes the picture of C. granti therein as correctly identified (i.e., Sands does not distinguish on gross snout morphology). Elwin described Corydoras arcuatus in 1939, from aquarium imports routed via Germany stating that at least one specimen was said to originate Brazil, Teffe and about the 50th species described although about a quarter of these are junior synonyms today. The details within the description are worth looking at a little more closely.

The description of Corydoras arcuatus appears in the January 1939 edition of the Magazine of Natural History, which was published late in 1938. In it, Elwin notes an import into England from Germany of a clearly new species of Corydoras from an unknown location in South America. She then goes on to note that the description that follows is based on a specimen deposited in the British Museum of Natural History (BMNH). We then find the curious statement that "Since the type was described, a smaller specimen has been received for determination at the BMNH from Mr J. Paul Arnold of Hamburg [Germany]. The locality of this fish is given as Teffe[sic], Amazon, and notes on it are given below". Immediately below however is the description of the type, not Arnold's specimen, and the type is stated as 46mm SL.

Dutch congress of aquarium clubs in late 1937 or early 1938, Elwin (now 29) second from right.
Water Life Magazine 22 February 1938.

The description of Arnold's second specimen mentions 37mm SL, and slightly less high-bodied, slightly shorter headed fish that otherwise matches the first description. Thanks are then noted to Mr L. C. Mandeville (Elwin's husband, whom she had married in 1936) for photographing the live specimen which is clearly what we now call C. granti. Knowing they kept, photographed and wrote about many fishes they kept at home, it is reasonable to assume this was a fish in their care. We all missed an opportunity to ask her - Margery Elwin died in Bristol at the age of 96 just a month after Steve's CSG article was published. The upshot of all this is that Elwin had both C. arcuatus and C. granti in her care and it is one of luck, its untimely demise, or perhaps she thought fit to deposit the largest specimen, at the BMNH that has led us on this most excellent adventure. Further specimens, and thus photographs would have been curtailed by the outbreak of World War II later that year.

I did consider with all this mention of second specimens and the fact that the type is not the photographed type, that somehow something got mixed up, the museum, and its fish collection was not unaffected but the unfolding of the War later in 1939. The paper records in the BMNH (Zoology Accessions Register: Fishes: 1937-1960: page 38) show that both the specimen was deposited in March 1939 (months after the description was published) and that this was the first specimen received (denoted by a pronounced dot in the 2nd column). However, the description mentions the specimen deposit at the BMNH being made the prior year. The reason for this gap in time remained a mystery I could not solve or indeed hypothesise a set of events beyond it just happened later. However, despite the irregularities surrounding a long-misidentified fish, it does not appear to be a case of mixed-up types for a single reason I can see. The specimen in the description is the exact match of 46mm in the description with the type and we must assume the photographed fish, while described in the paper as a type, was just another specimen the Mandevilles had a home - so an unpreserved paratype. This is further confirmed in the description of C. granti that the morphology matches the BMNH holotype.

BMNH specimens logged in 1939 include Elwin's C. arcuatus - the dot left of the species name indicates this is the first individual of this species in the BMNH collection. © BMNH and their trustees.

The other C. arcuatus at the museum were from a later, seemingly unrelated, donation in 1958. © BMNH and their trustees.

Outside of Europe, records show the fish was in the New York City Aquarium by at least 1939 and older American literature notes first importation in large quantities into Germany and the US around this time. An early common name of Tabatinga Catfish was used to describe "a tributary of the far reaches of the Amazon where it was collected". There are several Tabatingas in Brazil, but this can only be the Brazilian City of Tabatinga on the border with Colombia (wherein the city continues and is known as Leticia). A mainstay export spot, and somewhere Corydoras have been exported from as long as there have been exports. The initial collections were made by the noted collector, Rabaut.

The mysteries have been thoroughly unravelled and the misidentified corrected. Today, as we see species of Corydoras grouped by their lineages, this makes it easier to understand all of them as a whole and also will assist avoid such misidentifications in the future. In business, we talk about verticals (markets) and horizontals (capabilities); this translates to the cory world as the lineages (verticals) and lookalikes (those that share a pattern, or horizontals). As Corydoras is the name for the oldest, most basal lineage one, then these horizontals can be known by their "leader". Corydoras narcissus in this case. And so, the narcissus-group of Corydoradinae is now a well-defined horizontal and, as I've said, a mystery unravelled.

In closing, I wonder how many other famous names might be similarly affected. Some of the most commonly kept species such as C. sterbai, C. similis or C. panda all could be candidates for another excellent piece of taxonomic detective work. I look forward to reading about it!

With thanks to Steve Grant for advice, patience and corrections to the above.

For more on Elwin, see Water Life Magazine 1936-1958: Part 2. The Editor, Margery Graves Elwin.

There is further information on this species on the Cat-eLog page.

Back to Shane's World index.