It has always been rumoured that some snakes grow to sizes that exceed the 10 m record generally accepted as the authenticated maximum: this was for a Reticulated python Python reticulatus shot on Sulawesi in 1912. Numerous stories and anecdotes discuss Reticulated pythons and anacondas Eunectes murinus that far exceed this, with the most famous of these stories being Major Percy Fawcett's 19 m long anaconda that he claimed to have shot in the Brazilian Rio Abuna in 1907 [scene depicted in adjacent image]. Despite its immense length, Fawcett reported that this snake had a width of just 30 cm (which makes the tale rather unbelievable), and he even discussed trails in the swamps that 'support the statements of Indians and rubber pickers that the anaconda sometimes reaches an incredible size, altogether dwarfing that shot by me' (Heuvelmans 1995, p. 340). Despite these many stories, no-one has yet produced a snake that exceeds 10 m in total length (though I hope that - if such a specimen is ever found, it is not killed, but is restrained or captured or something.. and yes I do know that this is somewhat less easy than it sounds).
Perhaps less well known than these accounts of modern-day giants, however, are rumours and reports of fossil snake bones that have been estimated by some workers to have belonged to truly gargantuan reptiles...
As is always the case when several large-bodied members of a group are known, experts disagree over which of the several contenders is 'the biggest'. Among fossil snakes, the disagreement has centred over Gigantophis garstini of Upper Eocene Egypt and Libya*, first described by Charles William Andrews in 1901 (Andrews 1901, Hoffstetter 1961), and Palaeophis colossaeus from the Eocene of Mali, named by Jean-Claude Rage in 1983 (Rage 1983). Gigantophis is today regarded as a madtsoiid**: this is a group of controversial archaic snakes whose fossil record extends from the Upper Cretaceous to the late Pleistocene (though they only survived beyond the Eocene in Australia). Snake workers differ in their interpretation of where madtsoiids fit within snake phylogeny, but that's a debate that I want to avoid here (see Rieppel et al. 2002, Scanlon 2003, 2005, 2006, Scanlon & Lee 2000, 2002). Palaeophis belongs to the palaeophiids, a group of unusual marine snakes known from the Cretaceous to the Eocene. They also appear to be archaic and may be among the most basal snakes known: I don't want to say too much about them, as a blog article devoted to them alone is due to appear soon.
* It now seems that some of the Fayum deposits conventionally regarded as Upper Eocene are Lower Oligocene (Seiffert 2006). I do not know if this affects the age of Gigantophis.
** Incidentally, how are you supposed to pronounce this word? I say it 'mad-tee-soy-id'.
So how big were these snakes? Gigantophis has generally been given a length of about 9 m, but Rage (1983) suggested that Palaeophis colossaeus might have exceeded this and thus been the biggest snake of all time. Whether he meant that it exceeded 10 m isn't clear, but note that even these mega-huge snakes are either smaller than, or not much bigger than, the biggest, record-holding living snakes. Jason Head and David Polly recently used morphometric techniques to provide more thorough length estimates for Gigantophis: using the proportions of modern snakes as a template, their regression analysis of body length against vertebral size (this taxon is only known from isolated vertebrae) suggested a total length of 9.3-10.7 m 'indicating that this is the largest snake for which such an analysis can be performed' (Head & Polly 2004) [adjacent image of Gigantophis, to scale with a person, from here].
The authors didn't fail to make note of the fact that Gigantophis lived alongside a diverse assemblage of basal proboscideans, big hyracoids, amphibious artiodactyls and other taxa, and it is tempting to speculate that this immense snake preyed upon some of the larger mammals in this community. The Upper Eocene/Lower Oligocene of Fayum was also home to enigmatic giant flightless birds (Eremopezus), the bizarre ptolemaiids, early macroscelideans, big creodonts, archaic primates and many other neat taxa [image below shows Gigantophis eating a moerithere, from here. It is sometimes said in the literature that madtsoiids couldn't distend their jaws in the same manner as extant snakes, so I don't know if a feat like this was really possible].
As fossil snakes go, Gigantophis is reasonably well known, and deservedly so given its size. Little discussed, however, are a few other, far more obscure candidates for the title 'largest ever snake'. In a paper on Brazilian sebecosuchians, Zulma Gasparini and colleagues (1993) made a brief passing mention to 'huge booids up to 29 meters in length'. Cough splutter... 29 m?? They chose not to elaborate on this interesting observation, and only cited Adriana Albino's 1989 doctoral thesis, which I haven't seen. Also in 1993, Albino described the new boid Chubutophis grandis from the Eocene of Chubut Province, Argentina. Unfortunately C. grandis is known only from a partial body vertebra. However, it is especially big and a few features suggest that it's from a juvenile. Albino wrote 'Considering that this material is from a young specimen, its size is extraordinary and it is estimated that the adult would have reached greater dimensions than those observed in the largest snakes known up to the present, including Madtsoia and Gigantophis' (Albino 1993, p. 60). Elsewhere in the same paper, Albino noted that Madtsoia 'was probably 10 m long' (p. 63) and again made the point that the adult form of Chubutophis 'would have surpassed Madtsoia in length'. A 10 m length estimate for Madtsoia is interesting: one of the largest species in the genus, M. bai from the Eocene of Argentina has been estimated at 7-8 m (Smith 1990), but I haven't seen the 10 m estimate elsewhere.
I therefore wonder if C. grandis is the same thing as the '29 m long' boid referred to by Gasparini et al. (1993). In 1999 I asked Jean-Claude Rage about these giant snakes. He informed me of a 1991 popular article by Albino (note: a popular article and not a technical paper) in which it was suggested that the biggest ever snakes reached lengths of between 15 and 20 m. In this article Albino discussed: (1) a juvenile snake from Eocene Argentina estimated to have been 5-7 m long, and in adult form 10-12 m long; and (2) another Argentinean snake known only from an incomplete vertebra, estimated to belong to a snake with a head 70 cm long, and a total length of 15-20 m (Rage suggested that the 29 m length given by Gasparini et al. (1993) might be a typo). Unfortunately the age of the second form is unknown. It seems that the first snake Rage referred to was Chubutophis but, if this is true, then Albino wrote of two different really huge snakes. I therefore confess to being confused and I suppose the only person who could sort it out would be Dr Albino: despite attempts a while back to contact her though, I was unsuccessful. I asked Jason Head what he thought about the possibility of a 29 m long fossil snake. His answer was that such a creature was highly unlikely to have existed, given various constraints (though his actual language was a bit more colourful: we were in a bar at the time) [adjacent image shows pretend anaconda from the god-awful movie of the same name. Didn't it star Jennifer Lopez?].
For the record, I agree with Jason that snakes exceeding 15 m in length almost certainly never existed, but I still want to know why a qualified fossil snake expert would say, on more than one occasion, that they did. At least one noted expert on fossil snakes is a regular visitor here at Tetrapod Zoology. At the risk of putting him on the spot, any more information would be gratefully received.
Finally, obviously this isn't the 'new mammal diversity' article I've been plugging lately. Like cassowaries kick ass, this post has appeared out of lack of time and desperation to post something new: it was a bit of old text that only took a short while to update. Dammit, if only I didn't have to work.
Refs - -
Albino, A. M. 1991. Las serpientes del Santacrucense y Friasense de Argentina. Ameghiniana 28 (3-4), 402.
- . 1993. Snakes from the Paleocene and Eocene of Patagonia (Argentina): paleoecology and coevolution with mammals. Historical Biology 7, 51-69.
- . 2000. New record of snakes from the Cretaceous of Patagonia (Argentina). Geodiversitas 22, 247-253.
Andrews, C. W. 1901. Preliminary note on some recently discovered extinct vertebrates from Egypt (Part II). Geological Magazine (Dec. 4) 8, 434-444.
Gasparini, Z. 1993. New Tertiary sebecosuchians (Crocodylomorpha) from South America: phylogenetic implications. Historical Biology 7, 1-19.
Head, J. & Polly, D. 2004. They might be giants: morphometric methods for reconstructing body size in the world's largest snakes. Journal of Vertebrate Paleontology 24 (Supp. 3), 68A-69A.
Heuvelmans, B. 1995. On the Track of Unknown Animals. Kegan Paul International, London.
Hoffstetter, R. 1961. Nouvelles récoltes de serpents fossils dans l'Éocène supérieur du désert Libyque. Bulletin du Muséum National d'Histoire Naturelle, 2e Serie 33 (3), 326-331.
Rage, J.-C. 1983. Palaeophis colossaeus nov. sp. from the Eocene of Mali, with remarks on the genus problem within the Palaeophinae. Comptes Rendu de l'Academie des Sciences, Paris, Serie II 296, 1741-1744.
Rieppel, O., Kluge, A. G. & Zaher, H. 2002. Testing the phylogenetic relationships of the Pleistocene snake Wonambi naracoortensis Smith. Journal of Vertebrate Paleontology 22, 812-829.
Scanlon, J. D. 2003. The basicranial morphology of madtsoiid snakes (Squamata, Ophidia) and the earliest Alethinophidia (Serpentes). Journal of Vertebrate Paleontology 23, 971-976.
- . 2005. Cranial morphology of the Plio-Pleistocene giant madtsoiid snake Wonambi naracoortensis. Acta Palaeontologica Polonica 50, 139-180.
- . 2006. Skull of the large non-macrostomatan snake Yurlunggur from the Australian Oligo-Miocene. Nature 439, 839-842.
- . & Lee, M. S. Y. 2000. The Pleistocene serpent Wonambi and the early evolution of snakes. Nature 403, 416-420.
- . & Lee, M. S. Y. 2002. Varanoid-like dentition in primitive snakes (Madtsoiidae).
Journal of Herpetology 36, 100-106.
Seiffert, E. R. 2006. Revised age estimates from the later Paleogene mammal faunas of Egypt and Oman. Proceedings of the National Academy of Sciences 103, 5000-5005.
Smith, M. J. 1990. Wonambi naracoortensis. In Rich, P. V. & van Tets, G. F. (eds) Kadimaka. Extinct Vertebrates of Australia. Princeton University Press (Princeton, New Jersey), pp. 156-159.
More like this
My best guess: "madt" rhymes with German stadt, + SOI-ids. So, three syllables -- madt-SOI-ids.
Looks like Madts is a Danish variant on the surname Matthews.
(I don't know beans about zoology, but I try to contribute where I can. :-) )
Very interesting topic! I still remember the time when Gigantophis was estimated to be around 20m, so I wouldn´t be surprised if those mysterious monster-snakes will also shrink in the future.
I have in fact problems to believe in a 29m long snake. Such a giant would weigh 33 times the weight of a already monstrous 9m snake. Even if you keep in mind that pyhtons are for example much slimer than anacondas, this thing would still weigh at minimum 3.500kg or so, or with anaconda-proportions probably more in the range of 6000kg+.
By the way, there was once a huge pyhton in the Bronx Zoo named Colossus. When it died on a tuberculosous spine-deformation, it was more than 9m (I have the correct length in a book) long, and untill its death if had grown more than 20cm per year in the Zoo. If it had become older, Colossus had perhaps become a new world record holder.
While a bit anecdotal, I'll add my 2 cents worth. Zoo herpetologists seem to agree that the biggest extant snakes were historical records, which can be interpreted in 2 ways: the age structure of wild snake populations are more skewed toward younger snakes, and/or that limitations on the wild animal trade have dampened collecting effort. When I was working in and out of Indonesia in the late 70s I never failed to encounter newspaper articles about pythons eating people. Macabre. Also, it wasn't uncommon on morning rounds for keepers to find a mammal exhibit replaced with a 'bloatatious python'. After eating the binturong or muntjac they were too big to squeeze out of the cage.
What constraits actually limits the max size of snakes?
I only can think of one size related problem and that would be suffocation on their own weight putting pressure to the lungs.
But theres an obvious sollution to this...become aquatic.
Ok this is just amateurish speculation but I would be really interested on the constraints that limit the max size on the snakes.
Why "the"? It's just the island today called by its undistorted name Sulawesi.
On the pronunciation of "madtsoiid" I'm with HP: matt-tsoi-id. You're not supposed to just invent a vowel. :-)
Incidentally, how are you supposed to pronounce this word? I say it 'mad-tee-soy-id'
MADT-soh-EYE-ihd. As noted, the "madt" is like "mat" but with a d-ish sound in it, as if you were saying "Madsen" but stopped before getting through the "s."
Well, according to Wikipedia (take that for what it's worth), Madtsoia is from Tehuelche mad 'valley' plus tsoi 'cow' and is a reference to the type locality, Cañadón Vaca. Accordingly, "madtsoiid" should be pronounced mad-TSOY-id, though I imagine mad-SOY-id would be OK for an English speaker who isn't used to pronouncing a [ts] affricate at the beginning of a syllable.
Hi Darren, I guess you have put me on the spot. I can tell you how I pronounce 'madtsoiid', but let's go back to the source... An important reference you didn't cite is
Simpson, G.G. 1933. A new fossil snake from the Notostylops Beds of Patagonia. Bulletin of the American Museum of Natural History 67: 1-22
which is freely available here. Simpson named Madtsoia bai as a new genus and species of boid, using words from the local Tehuelche language: mad and tsoi as a tranlation of the locality name Cañadon Vaca ('cow valley'), and bai, 'grandfather'. I doubt that Tehuelche or many other languages (I won't say any) naturally include the cluster 'dts', but seeing it as two separate words run together may make it easier to say. 'ts' is just the 'hard s' often spelled with a 'c' in English and other languages (as in 'once'). I read the 'oi' as a diphthong, so pseudo-phonetically I make it 'mad-tsoy-a' and 'mad-tsoy-id'.
Simpson also came up with 9 m as an estimated length; the biggest piece of his type specimen was 40 vertebrae in articulation, with ribs, which is the most ever found from a snake of that size. Andrews said '30 feet' for Gigantophis, which of course is just the same size.
The biggest Australian madtsoiids were maybe 6.5 m; the most complete skeleton is a subadult with 25 of the vertebrae articulated but a lot of the rest missing, so the estimate's pretty rough so far. We've actually got a python from the Pliocene that was bigger than that, possibly over 9 m.
As for the maximum size snakes could attain... well, having seen various species of goannas (Varanus), who would ever have predicted V. priscus ("Megalania")? I happen to possess a juvenile priscus vertebra that's way bigger than anything from komodoensis, just as Chubutophis is bigger than Madtsoia; and if that were all that we had of priscus I would understand your skepticism. I'm loath to estimate the biggest a snake could get: but I'd like to see some adult Chubutophis bones one day.
Imagine my disappointment at finding, today, that the 49-foot-long python reported in 2004 in Sumatra was really less than half that long.
On another subject, I can just about swallow monster birds as tetrapods -- two isn't so far from four -- but I have to draw the line somewhere short of snakes. Zero is a fundamentally different number from four.
Many thanks to all for their comments. I can rest assured that I'll no longer sound silly when I (properly) pronounce madtsoiid :)
Nathan: you're referring to Fragrant Flower. Every now and again someone tells me about this immense Sumatran snake that well exceeded the 10 m limit. How disappointed they are.
And on the whole 'tetrapod' issue. You're joking I'm sure :) Tetrapoda is a clade, defined ancestrally by the presence of four limbs, but with many lineages that have since modified this pattern (in other words, tetrapods are all those vertebrates that possess four limbs, or descend from ancestors that had four limbs).
Wrong.
It does occur in English, but only at the ends of words, as in "lists" or "oats".
Take the English ch. It is t + sh. Now try that with s instead of sh...
Well, yes and no. For many speakers, "once" does end with a [ts] affricate*, but that's not because it's spelled with a c; it's because speakers automatically insert a [t] sound into what would otherwise be an [ns] cluster. As a concrete example, for these speakers, "prince" is pronounced identically to "prints". This effect arises from slight differences in the relative timing of the movements of the tongue tip and of the opening to the nasal cavity.
That said, it is true that in the orthographies of Baltic and Slavic languages that use the Latin alphabet (Latvian, Lithuanian, Polish, Czech, Slovak, Slovene, Croatian, etc.), and in the orthographies of some Native American languages, the letter c stands for the affricate [ts].
*An affricate is a sound that starts with the air completely stopped off, then releases in to a fricative sound (fricatives include sounds like like [f], [s], the "th" in "think", and the "sh" in "shirt").
Off-topic, but:
So, Darren, when will we see your obligatory post on the new Nessie footage? Your fans demand it.
If Fawcetts anaconda was 30 cm thich, it definitely was nowhere near 19 meters long. I've seen an anaconda that was thicker than that, and while it was quite a monster it was no record-breaker. It was lying coiled up in a thicket so I only got a half-decent picture of it and couldn't judge just how long it was, but the locals who knew it well claimed that it was about 7 meters.
About those immense trails, that anaconda was quite easy to locate, exactly because it had made a very distinct trail where it used to "launch" itself into the river, so all one had to do was to back-track along it. However the track was wider than the snake was thick, so it is not straightforward to judge the size of the snake from the track.
Dear Dr. Naish, I visited your blog because John Scanlon told me that you tried to contact me to discuss size estimates for Chubutophis, without success.
I share your skepticism about the existence of extant snakes more than 10 m long, but, after the reading of your topic Stupidily large snakes. there are some comments I can made.
First: The mention of Gasparini (1993) about a 'huge booids up to 29 meters in length' was obviously a mistake of Gasparini´s publication. I never said that in my doctoral Thesis.
Second: There are some mistakes in the references:
1) - . 1993. Snakes from the Paleocene and Eocene of Patagonia (Argentina): paleoecology and coevolution with mammals. Historical Biology 7, 71-69.
The correct pages are 51-69.
2) - Albino, A. M. 1991. Las serpientes del Santacrucense y Friasense de Argentina. Ameghiniana 29 (3-4), 402.
The number of the article is 28 (instead 29), but this resume is not the popular article where I discussed the size of Tertiary southamerican snakes. The article I suppose you wanted to cite is:
1991. Serpientes gigantes en la Patagonia. Ciencia Hoy, 3(14): 58-63.
Ciencia Hoy is the most well-reputed publication on divulgative science in Argentina since the ´80 decade.
In this article I discussed: (1) a juvenile snake from Eocene Argentina estimated to have been 5-7 m long, and in adult form 10-12 m long. This vertebra was posteriorly described as Chubutophis in Albino (1993). (2) another Argentinean snake known only from an incomplete vertebra, estimated to belong to a snake with a head 70 cm long, and a total length of 15-20 m. This last remain is an incomplete centrum with the left prezygapophysis of an adult specimen (there is a photo in natural size in the article and a filmation of the remain in a documental film directed by Jorge Prelorán). I never publicated this last remain in a scientific paper because stratigraphic age and locality are unknown, although I could established that it is from early Tertiary levels in south Chubut province, Patagonia. Because the remain is extremely bad preserved, I cannot determine it as a boid or a madtsoiid snake.
I estimated the size of these snakes only on the basis of the length of the vertebral centrum because geometric models, as that used by Head and Polly (2004), are recent. I took into account the length of the vertebral centrum of a mid-trunk vertebra in an specimen of boid with known total length (and known head length), and the number of vertebrae that integrated the vertebral column. So, I extrapolated this information for fossil snakes considering that boid species aproximately have 300-400 vertebrae. Nevertheless, it is important to assume that we do not know how many vertebra form a vertebral column of a fossil snake!!.
Third: In addition, you said that a 10 m length estimate for Madtsoia by me in 1993´s paper is interesting because one of the largest species in the genus, M. bai from the Eocene of Argentina has been estimated at 7-8 m (Smith 1990), but you haven't seen the 10 m estimate elsewhere. Well, I have the oportunity to take measurements on two articulated vertebrae of a Madtsoia bai holotype cast in the Museo Argentino de Ciencias Naturales (you know.the holotype collected by Simpson in the Chubut province of Argentina was deposited in an institution outside my country when Argentine´s laws about fossil exportation were permisive). I could estimated a 10 m long Madtsoia specimen on the basis of the measures taken on this cast, following the same method I described above.
I hope my comments can clarify the question about largest fossil specimens of snakes. If you have another doubt you can write me an e-mail (unfortunately I do not have the habit of blog reading).
I also hope you can excuse my extremely horrible English.
Adriana
Dr Albino: many, many thanks for visiting and leaving the comments and corrections, it is very much appreciated. I am pleased to see that I remain unable to write down numbers correctly :)
I tried to think of a vertebrate with a negative number of feet to exemplify the incremental danger of entertaining snakes as tetrapods, but none come immediately to mind. Cf. Melville on whether to count whales among fishes, and Tolkein on dragons as hypertrophied worms. I don't know which kingdom Lovelock put Gaia in... one could argue for Archaea, although fossil evidence is sparse, and extant species are much reduced, by comparison.
And, yes, of course I meant to write "terror birds" rather than "monster birds". When you have children you will begin to make mistakes too.
It is too late for me, my friend (though I'm not yet in the plural of 'child').
Just an additonal question for Darren and all the others who read this: Some friends of me once told me that they have seen a discovery channel episode about the alleged track of a monstrous anaconda which was found in the jungle of South America, and which was estimated to come from a snake of 30-40m. To make my positon about this clear, I really don´t believe that there are any anacondas only close to this size out there, but those friends are very hard to convince. The note of Tommy Tyrberg that tracks of such snacks are much wider than the snakes themselves is very interesting. Does anybody know about this episode of discovery channel or can me give further prooves that this track (which was shown in the episode) can not belong to a 40m long snake ?
Darren
There was a ultimate snake programme recently on National Geographic that mentioned a 26ft anaconda that was captured in Venezuela by the police, would this be a world record size for this species?
Unfortunately despite the film lasting 1-2 minutes there was no verification of the snake's actual size.
I have seen snake trails in the Amazon, but I don't imagine that I can can correlate width to length.
Hi Paul, thanks for your comment. There are several claimed records for anacondas round about the 10 m record generally accepted for reticulated pythons, but which are accurate and which not is difficult to gauge. A 10.4 m long anaconda was supposedly shot in British Guiana by the author, museum director and philanthropist Vincent Roth, and a 10.26 m specimen was reported from the Guaviare River by Colombia University's Dr Frederick Medem. The most oft-mentioned giant is the 11.4 m long specimen shot in 1944 in Colombia. The story goes that the discoverers (a prospecting party led by geologist Roberto Lamon) shot it dead, then wandered off and ate lunch. When they returned to photograph the snake, it had vanished.
Given that some of these specimens were reported by reputable individuals, I'm not sure why that reticulated python from Sulawesi is so often listed as 'the' biggest modern snake, especially given that anacondas are heavier for their length. Maybe it's because the Sulawesi retic was (apparently) measured very accurately (by civil engineers using surveying tape).
Anyway.. no, the 26ft Venezuelan anaconda is clearly not the largest that has been reported.
Sources - -
Carwardine, M. 1995. The Guinness Book of Animal Records. Guinness Publishing, Enfield (Middlesex).
Mattison, C. 1998. The Encyclopedia of Snakes. Blandford, London.
A comment on Adrianna's vertebral counts used in estimating body size in extinct fossil snakes (and a shameless plug for a recent pub'): Gigantism in snakes results in fewer than expected vertebrae relative to body size in multiple lineages. So, estimates of body size in fossils should use conservative estimates of vertebral #.
See:
Head, J. J., and P. D. Polly. 2007. Dissociation of somatic maturity from segmentation drives gigantism in snakes. Biology Letters, 3(3):296-298. doi:10.1098/rsbl.2007.0069.
it is a very intaresting article for people who likes traveling it is exelent.
Whether he meant that it exceeded 10 m isn't clear, but note that even these mega-huge snakes are either smaller than, or not much bigger than, the biggest living snake.
If we assume that the giant snake fossils represent average size, then they are much longer on average than Python reticulatus which probably averages between 5 and 6 meters (for females). It's a big pet peeve of mine when books say "Anacondas are 20 feet long" even though such lengths are actually rather unusual.
Is the post on Palaeophis and kin still on your list of plans?
Yes - stay tuned.
Gigantophis fits right in inhabiting the exotic locale of the Fayum in the Oligocene, where such flamboyant animals as Arsinoitherium. Paleomastodon, and various extinct primates are found. BTW, I hope that isn't true about Gigantophis not being able to open it's mouth as wide as modern snakes, because that picture of one swallowing a Moeritherium is really cool!
With the new Columbian find of Titanoboa cerrejonensis the author and Jason Head may want to reconsider their statements regarding snakes over 15m never existing.
Titanoboa was "only" 13 m long.
Never believe anything anyone tells you in a bar......
is that snake real
Which one?
A long time ago I had a thump in my heart when I unexpectadly crossed a 10 inch snake and keep curious about them. I enjoyed your article about Titanoboa.
Who really knows what creatures are still to be discovered in the depths of the Amazon Jungle? Giants snakes or other yet undiscovered creatures.
What is interesting is that though Colonel Fawcett was ridiculed at the time for his report of seeing a giiant snake and the double-nosed dog, it seems that he may have been telling the truth after all.