c. comma Variation B
The distribution range covers the KwaZuluNatal Province of South Africa, Mozambique and Madagascar. The shape of the shell is oval, often callous. (Massier has in his collection specimens of this oval, callous form originating from mainland Tanzania as well as Zanzibar.) The labral margin is rounded and rather callussed. The columellar margin is often rather callussed. The marginal spotting is often distinct along the upper labral margin with some spotting rarely on the columellar margin and the base always unspotted. The ground colour is saturate purple with distinct darker transverse banding. The dorsal colour is very dark brown, forming small lacunae.

Merlin is of the opinion that specimens in his collection of Variation B from Mozambique and Madagascar, are similar to Variation A and that these are often confused with e. francescoi : "Dans la forme B toutes celles que nous avons du Mozambique et de Madagascar sont semblables à la forme A. Il semblerait qu'il y ait souvent confusion avec francescoi.

Similar-looking specimens are also found on northern Unguja Island, Zanzibar, Tanzania. A sample of three specimens show lengths ranging from 14.2 - 18.8mm. Only one of these specimens shows spotting on both margins. Two further specimens from Shanzu Village and Diani Reef in Kenya also bear resemblance. All of the specimens except two have a dark base. The dark base is not seen in the Tanga Bay specimens. [Plate 11]

The further discussion of this as a definite form of c. comma is inhibited by the lack of a sufficient number of specimens from these localities. The name "tanga" might well be invalid because the form could extend beyond Tanga Bay.

This should be considered an anecdotal exercise to identify a specimen with prominent differing features. However, if a case can be made out that this is indeed a specimen of e. francescoi form mainland, then it must be considered the earliest attempt to name the form.

Specimens resembling c. abaliena in size and colour originate from Zanzibar, Tanzania and Reunion. These are adult shells. The length ranges between 11.5 mm - 16.2 mm. They differ from c. abaliena in that they are significantly more slender, do not have callous margins and the lacunae are smaller and more typical of c. comma. [PLATE 12]

Is this a variation of c. comma? Or an intermediate between c. comma and c. abaliena? Lorenz accepts the sharing of the same geographical region of these two ssp. just as he acknowledges the co-existence of e. esontropia and e. cribellum and by cumingii cumingii (SOWERBY I 1832) and c. astarayi (SCHILDER 1971).

In the paratypes, the degree of marginal callosity varies considerably. There are slender, thin walled shells as well as very broad callussed specimens. The degree of marginal spotting varies too... the dorsal coat varies in the size and number of lacunae. In nearly all specimens studied, there is no pattern along the labral fourth of the shell."

From his earlier commentary on pelliserpentis, it appears as though Lorenz favoured assigning francescoi to c. comma.
'In Madagascar and Mozambique, an interesting variation of Cribrarula cribraria comma (PERRY 1811) occurs (variation B), which differs from typical specimens (variation A) by distinct spotting along the dorsal edge of the labral margin. This spotting, however, is not present along the columellar margin or on the base. This feature is a fundamental difference to the Mascarenan Cribrarula esontropia DUCLOS, which has lead to confusion in the past, e.g. by myself (see Lorenz 1998). The true Cribrarula esontropia (including its subspecies cribellum) seems really endemic to Mauritius and La Reunion, while all records of the species from other areas in the Indo Pacific and the Persian Gulf seem to be based upon exceptionally well spotted specimens of cribraria comma, cribraria ganteri and the new taxon described herein, Cribrarula pellisserpentis sp. nov."

Lorenz later revised this and assigned francescoi to esontropia rather than cribraria comma :
"In e. francescoi ... the collumelar teeth are coarser and less numerous than in all East African species of Cribrarula. This is the main difference also from the East African cribraria comma (PERRY 1811)... in which the labral spotting is rarely found labrally, but is usually absent. Also c. comma has both lobes of the pattern. ... The closer conchological resemblance allows a placement of e. francescoi as subspecies of esontropia rather than cribraria."

In summary, Lorenz writes :
"The recently discovered francescoi from Madagascar differs from its relatives by the absence of the labral lobe of pattern, by stronger dentition and by reduction of the marginal spotting along the left side."

However, spotting on the labral margin of Madagascar specimens, can be intensely dense and extensive to the point of having an appearance of a solid line. [Plate 14]

Amongst malacologists there is disagreement about such a placement. Some would favour it to be a ssp. of cribraria, i.e. cribraria francescoi and there is also the opinion that it should be considered a form of c. comma, i.e. c. comma form francescoi.

The latter opinion is that of Massier who motivates his arguement as follows:
"According to Lorenz's map in his new book cribraria comma should not be found further south than Central Mozambique. Well, I have some typical ones from southern Mozambique and northern Natal. Most shells I have from the mainland are in-betweens of c. comma and franciscoi. Only a few are real franciscoi. This is also an indication that francescoi should rather belong to cribraria and not esontropia. The typical flat and squat shape of esontropia is not found in francescoi. I have never seen in francescoi the heavy marginal spotting of esontropia which even reaches the base. On the other hand I have seen many cribraria comma with the same spotting as francescoi and there are also francescoi with no spotting but all the other features.

I can also conclude that francescoi should be called cribraria francescoi and not esontropia francescoi. Seeing that there are so many integrades with cribraria comma and that the geographical distribution overlaps I would call francescoi a form of c. comma and not a subspecies."

Massier later added additional comments:
"I cannot detect the dark dorsal banding in francescoi mentioned by Lorenz in any of my specimens. My tooth count resulted to nothing. In francescoi I counted between 11 - 15 columellar teeth and in comma 12 - 15. This is also supported by the teeth statistics in your work. So I cannot support Lorenz's statement: 'In e. francescoi the columellar teeth are coarser and less numerous than in all East African species of Cribrarula.' I have to state that I do not have enough specimens in my collection to justify a proper statistic with regards to the teeth."

The authorship of francescoi is disputed by Merlin and Philippe Quiquandon, France. They claim to have identified it as a ssp. of Cribrarula in advance of Lorenz's publication of his description. According to Merlin & Quiquandon they showed specimens of these shells that hailed from Madagascar, to Lorenz at the Paris Seashells Show 2002. Lorenz identified them as e. francescoi. The French then published their notes and photographs of specimens in Xenophora. They identified the specimens as e. francescoi but posed the question whether the ssp. should not rather be associated with cribraria.

In an article in a later edition of Xenophora, Loïc Limpalaër endorses the Merlin & Quiquandon claim to have identified francescoi first: "As both the authors [Merlin & Quiquandon] explain they preceded the publication of this taxon by Felix Lorenz by a few weeks". And: "It was inadvertently christened in our columns of [Xenophoria] no 99: Cribrarula esontropia francescoi...".

e. francescoi form mainland and a possible variation thereof
Lorenz has yet to publish specific characteristics for what would constiutute e. francescoi form mainland. He does provide a photograph of a specimen on Plate 34:6 in "New Worldwide Cowries".

Xenophora edition no 102 has photographs of specimens from Mozambique and Madagascar bearing the identification Cribrarula cribraria francescoi. The Madagascar specimen meets Lorenz's description of e. francescoi and the Mozambique specimen would then qualify as Lorenz's e. francescoi form mainland.

Xenophora edition no 99 featured a photograph of a specimen from KwazuluNatal Province, South Africa which the authors labeled as c. comma. Because of the largely absent labral pattern and the spotting on both the labral and columellar margins, this specimen would also match Lorenz's e. francescoi form mainland.

Based on the Lorenz and Xenophora photographs and with specific reference to my own collection of mainland forms (obtained from and verified as such by Lorenz and other sources), I wish to present the following as a discussion of and description for e. francescoi form mainland:

14 Specimens of e. francescoi and 34 specimens of e. francescoi form mainland were available for study. [Plate 15, Plate 16, Plate 17] I am indebted to Jean-Claude Merlin, France, for detailed notes of the e. francescoi form mainland specimens in his collection. Merlin concurs with most of my own observations.

Following Lorenz's guideline, e. francescoi is restricted to Madagascar and e. francescoi form mainland is a relative found in Mozambique and the KwaZuluNatal and Eastern Cape Provinces, South Africa.

Shape
When comparing e. francescoi with e. francescoi form mainland, the immediate impression is that the South African and Mozambican specimens are not as slender as those from Madagascar. In fact, the mainland specimens appear globose and frequently more callussed and especially so amongst those from the Eastern Cape Province. (e. francescoi form mainland has a strong resemblance to exmouthensis. This is not the case with e. francescoi.)

Massier concurs : "I agree with your statement that on average the mainland francescoi is more bulbous than their counterparts from Madagascar, but in-betweens do exist. Merlin also states this fact."

Dorsal coat colour
There is also a difference in the dorsal coat colour with the mainland specimens ranging from chestnut to bright brown and those from Madagascar being deep brown.

Dorsal base colour
The dorsal base colour of e. francescoi show the distinctive green-purple hue whereas the mainland specimens have a more subdued hue that varies from cream-white to green.

Transverse banding
The embryonic banding across the dorsum is visible but not as obvious in the mainland specimens.

Lacunae pattern
There is substantially less overlap of lacunae and in some specimens there is no overlap.

Labral pattern
The labral pattern is often completely absent but more generally greatly reduced.

Marginal spotting
The spotting on the labral margin varies from sparse and scattered to extensive and stretching the length of the entire margin. Spotting on the columellar margin is infrequent and when present, restricted to the anterior extremity.

Teeth
The labral teeth of e. francescoi are short and fine but the mainland specimens have teeth that are short and coarse. There is a significant disparity in the relation of labral teeth to columellar teeth: the averages of 14 specimens of e. francescoi specimens are 13.50:14.00 whereas the averages of 34 mainland specimens are 14.06:13.94. (Merlin's figures for the averages of 7 specimens of the mainland form are 14.29:13.86)

Intermediates
Merlin also identifies intermediates between c. comma and e. francescoi form mainland. His specimens hail from Richards Bay and Park Rynie, both in the KwaZuluNatal Province. Richards Bay lies further north and is not far from the Mozambican border. The most significant difference is that the interior colour is predominantly pale compared with the generally dark interior colour of the true mainland specimens.

General comment
I favour a more descriptive naming of e. francescoi form mainland. Because of its prominent and eye-catching differences inclusive of the shell shape, colour, lacunae and teeth, I propose e. francescoi form robustus.

The most southern specimens of e. francescoi form mainland
Until late 2003 the most southern locality of specimens of e. francescoi form mainland in my collection was Park Rynie in the KwazuluNatal Province. I then acquired a collection 14 specimens from Lorenz that were beach collected in a more southern region stretching from Port Shepstone to Port Edward in the KwaZuluNatal Province. The shells were collected by Nolan Webb during the years 1972 to 1983.

With some exceptions, the specimens are of poor condition having lost most of the colour and lacunae pattern. The embryonic banding can be seen and the labral margin spotting is preserved. In most of the specimens the labral margin spotting is extensive, reaching from the posterior to the anterior extremities. Several specimens show very bold spotting. [Plate 18, Plate 19]

Massier has specimens from the same region in his collection that differ as regards the marginal spotting:
"I do have some shells from Mzamba (Northern border of the Transkei) which show all the francescoi features but no marginal spotting at all. All my francescoi (or c. comma) from Southern Natal or the Trankei are not live-taken. Often the dorsal pattern is very much reduced, but the marginal spotting, if present, is still feintly visible."

The shape and density of the lacunae and their overlapping are more reminiscent of e. francescoi from Madagascar. There is a far greater degree of lacunae overlapping than in the other e. francescoi form mainland specimens. In some specimens the netting pattern is vaguely similar to that of pelliserpentis. The three smallest specimens in the collection show heavy callussed margins. [Plate 20]

The poor condition of the beached shells deters me from speculating whether this collection of specimens might constitute a valid variation of e. francescoi form mainland. On average they are, however, distinctly different from mainland specimens collected further north.

REFERENCE SOURCES

Hawaiian Shell News
New Series No. 292 - April, 1984 - Vol. XXXII No. 4

Liltved, W. R., 2000
Cowries and their relatives of South Africa
2nd Enlarged Edition
Seacomber Publications, South Africa
ISBN 0-908-42089-7

Limpalaër, L., 2003
"Retour sur Madagascar. 2eme Partie: Sur les traces d'esontropia"
Xenophora: Bulletin de l'Association Francaise de Conchyliologie
Number 102, April - June 2003
ISBN 0755-8198

Lorenz, F. and Hubert, A., 2000
A Guide To Worldwide Cowries
2nd Revised and Enlarged Edition
ConchBooks, Germany
ISBN 3-925919-25-2

Lorenz, F.
Another new species of Cypraeidae from Madagaskar
(Mollusca: Gastropoda, Cypraeidae.) Cribrarula pellisserpentis sp. Nov.
http://www.cowries.info/shell/coandro/pelli.html

Lorenz, F., 2002
New Worldwide Cowries
ConchBooks, Germany
ISBN 3-925919-59-7

Merlin, J.C. and Quiquandon, P., 2002
"Les Variations de Cribrarula ou 'en attendant Lorenz'"
Xenophora: Bulletin de l'Association Francaise de Conchyliologie
Number 99, July - September 2002
ISBN 0755-8198

Personal correspondence with Lorenz, F.

Personal correspondence with Massier, W.

Personal correspondence with Merlin, J.C.

CAPTIONS FOR PLATES

Plate 1: c. comma from various west Indian Ocean localities

Plate 2: c. comma from various west Indian Ocean localities

Plate 3: c. comma with labral margin spotting

Plate 4: c. comma with columellar margin spotting

Plate 5: Different shell shapes of c. comma

Plate 6: Different lacunae forms and patterns of c. comma

Plate 7: Labral margin patterns of c. comma

Plate 8: Dorsal coat colours of c. comma

Plate 9: Dorsal ground colours of c. comma

Plate 10: c. comma "form tanga"

Plate 11: Look-alikes of c. comma "form tanga"

Plate 12: cribraria comma cf cribraria abaliena

Plate 13: e. francescoi from Madagascar

Plate 14: e. francescoi from Madagascar

Plate 15: e. francescoi form mainland

Plate 16: e. francescoi form mainland

Plate 17: e. francescoi form mainland

Plate 18: The most southern specimens of e. francescoi form mainland

Plate 19: The most southern specimens of e. francescoi form mainland

Plate 20: The most southern specimens of e. francescoi form mainland