ON THE EMERGENCE OF MAMMALS
(WITH SOME OUTLOOK TO THE SCARCITY OF P/T SPHERULES)
Summary of the lecture at the Geonomy Scientific Committee of HAS, Sept. 7, 2000
A report on the activity of the Matter Evolution Subcommittee
Evolutionary stories are sometimes more interesting than their details. We see a competition between living organisms of different Bauplans, both having specific gains and handicaps, and can learn from the story, how one and another try to make use of the changing environment.
The competition between synapsids and diapsids is even more interesting, for two reasons. One is that the competition started 310 My ago, when the ancestors of the two lineages, still almost indistinguishable, started to be land animals and started to diverge. It was an interesting competition with turnovers. (It seems now that the competition is practically over...) But the second reason is that the story is our story. The synapsids is our own lineage; after a fashion we are still synapsids.
But we may learn something else too from the competition. During the second half of Permian or at the Permian/Triassic (P/T) boundary something catastrophal happened. Several authors believe that it was the biggest catastrophe of Life, at least in Phanerozoic. The highest number which I saw was 97% extinction at species level. But we do not know what happened!
Since it is more or less known that the later Cretaceous/Tertiary (C/T) catastrophe was caused a rather small asteroid, the Dinosaur Killer, one school of experts look for another impact. But an impact should leave some traces. There are 3 classical traces of an impact: crater, extraterrestrial matter and spherules. The crater of the Dinosaur Killer has been found (Chicxulub, and maybe another near the Almirante islands, both under sea), the foreign matter in C/T layers has been found, and e.g. shocked quartz produced in collisions has been found at least at 7 widespread localities. Permian was 4 times farther in the past, still it is strange that there is no crater, no foreign matter. Some years ago IGCP 384 was formed to look for widespread global spherules, mainly from P/T boundary. It seems that its closing publications are still in preparation, and they collected a few P/T spherules, but it seems that the chemical composition of the spherules, if measured, are difficult to be interpreted. So we do not yet know if there was a P/T big impact; and if was, we definitely do not know the type of the impactor, if existed.
Now, if we observe the history of terrestrial life through the P/T boundary, we can at least get constraints, what may have happened. And since end-Permian and early Triassic were turning points of mammal prehistory, mammalogenesis is a source of information for the P/T global event.
1. WHAT IS A MAMMAL?
Table 1 gives traditional mammal characteristics vs. reptile in the left column; right columnlist mammals who do not fulfil (only living mammals):
Offspring sucking milk
from mother's teat
Constant body temperature
Body covered by hair
instead of scales
Skin is rich in glands
7 cervic vertebrae
Aardvark, some Cetaceans
incisors, canines, molars
Mandible-skull art.: D-Sq.
Table 1: Several "distinctive mammal characters" and mammals not sharing them.
Obviously a definition based on a similar list cannot be correct from the viewpoint of mathematical logic. Different "mammalian characters" are correlated but not absolutely.
Traditional classification (only living mammals) is simple because it seems that only 3 groups have survived:
Atheria vs. Theria = Monotremata vs. Ditremata
Theria -> Metatheria = Marsupialia vs. Eutheria = Placentalia
Table 2: The scheme of living mammal groups
1) It is impossible to make logical and operative classification using more than one definitions of the same rank. What is the conclusion, if an animal fulfils 6 criteria from 11 (Edentata cohors = Xenarthra & Pholidota)?
2) Recent reptiles are anapsids or diapsids, while our reptile ancestors were synapsid. So some recent mammal/reptile criteria may be rather synapsid/diapsid ones.
3) Monotremata asundered probably very early from the mammal stock, even possible that before the reptile/mammal transition.
4) In spite of traditional taxonomy in some fundamental characters Edentata differ more from the common mammalians than Monotremata do.
5) Xenarthra fossils before Late Palaeocene are not certain; the candidates belong to the group Gondwanatherium, which is either Eutheria incertae sedis, or Multituberculata, or a still unclassified Atheria with multituberculate affinity (see later).
6) Therefore until more fossils are found from Mesozoic, we cannot decide if Edentata are true Eutheria, or acquired placenta independently. Therefore we do not know if their peculiarities are from old independent origin, or they are synapomorphies inside Eutheria.
2. THE CHRONOLOGY OF TRANSITION
In this Table for convenience we give absolute ages. However the primary data are layers. Our numbers are based on a(n older) Calibration 1
Table 3: Calibration 1
Quite new literature pushes back J/C by cca. 10 Ma, and T/J and P/T by cca. 20 Ma (Calibration 2, later). Both calibrations come from radioactive dating.
In the following Table "reptiles" (Therapsida) are denoted by italics.
First appears in
At Ma BP
(on Eutheria degree)
Premolar vs. molar
Trigonal cusp pattern
Angulare moves to ear
Monotremata & Eutheria
Malleus & incus
7 cervic vertebrae
Cervic ribs absent
* Monotremes do not fulfil the criterion.
+ Edentates do not fulfil the criterion.
& Metatheres do not fulfil the criterion.
Table 4: A list of eutherian properties, with first occurrence. Italics denote "reptiles".
1) For a more detailed Table with 47 characters see , with references.
2) Apparently different criteria appeared at different times, this should prevent classification with multiple criteria.
3) Most "mammalian" characters appeared first in "reptiles".
4) Synapside end-"reptiles" (tritheledonts and tritylodonts) were already fairly "mammalian".
The time of the "reptile"-"mammal" transition depends on definition, somewhat arbitrary. For the "first mammal" serious propositions in these years are:
1) Eozostrodon, Uppermost Triassic/Basal Jurassic. Fairly complete skeletons.
2) Sinoconodon. 3-5 Ma earlier, partial skull. Still multiphyodont.
3) Adelobasileus cromptoni. Upper Triassic, from a layer gap. Partial skull.
3. MAMMAL TAXONOMY AND BRANCHING TIMES
In this Chapter times are meant Ma, BP. Normal font numbers mean consensus from fossils, italic fonts molecular clock, according to . In this Chapter we use the more contemporary Calibration 2:
Table 5: Calibration 2
Obviously the chronologies should be unified which we will do in the near future.
For the 3 recent mammal groups Monotremata (Atheria) and Theria asundered at cca. 165 Ma, while the Theria stock branched to Metatheria and Eutheria cca. 135 (173) My ago. (There exists a molecular date of independent Monotremata lineage from 134 Ma, but there are problems with that date. See Ref. 1.) From the extinct groups Multituberculata is the most important. It asundered from other Mammalia cca. the same time as Monotremata, or before, and died out in Eocene (?).
For living Eutheria the first branching was Edentata/Epitheria, 120 (129) Ma BP. Epitheria (i.e. all not Edentata) gave life first to 6 lineages:
1) Rodents & Lagomorpha 100 Ma.
2) Ferae 100 (92) Ma.
3) Insectivora 100 Ma.
4) Archonta 100 Ma.
5) Cete & Artiodactyla 100 (110) Ma.
6) Mesaxonia 100 Ma.
Later further branchings:
Rodents/Lagomorpha 70 Ma.
Archonta: Chiroptera, Dermoptera, Scandentia, Primates 75 (85) Ma.
Cete/Artiodactylia 60 (58) Ma.
Mesaxonia into Perissodactylia and Elephants + Sirenia 60 Ma.
In molecular chronology the Meta/Eutheria split is Mid-Jurassic, as old as traditional view thinks the Monotremata/Theria split. Generally molecular split times are older: lineages started to diverge when small opressed animals still did not differ too much in external appearance .
4. LATE PERMIAN AND P/T CIRCUMSTANCES
While the transition happened cca. in Upper Triassic, the cynodont synapsids appeared from Middle Permian and thenceforth the mammalian characters continuously increased. Since lot of guesses exist for the cause(s) of P/T extinction or Upper Permian troubles, we must look at the situation.
In Upper Permian our ancestors (synapsids) are the dominant land tetrapodes. In Uppermost Triassic we shrink, in Jurassic all synapsids become secondary compared to diapsids, and this remains so until K/T. What happened?
Cooling in Middle or Upper Permian. Glaciation. Anoxy in Upper Permian, minimal level 0.4 PAL in Middle Triassic, 1 PAL in Upper Jurassic, >1 PAL in Cretaceous.
First dwarving just after P/T, second from Upper Triassic. Global biocrisis at P/T.
P/T spherules exist (Miono + 1 spherule from Kashmir). However Miono spherules are not interpreted from impact.
Thermodynamic suggestion: In Upper Permian cynodonts are already warm-booded (degree unknown), hairy animals with heterodont, occluding dentition (intensive feeding) and double circulation. Cooling triggered the evolution of an efficient animal group, with high energy flux flowing in and out. Then oxygen level goes down, temperature up. Traditional diapsids return from the equator. They do not need too much oxygen.
From this point the synapsid/cynodont trend is a cul-de-sac. Oxygen deficiency forces small size (larger specific surface), heat loss is not a problem.
More details in  and in citations therein.
5. FROM CYNODONTS TO US
A complete cladogram is still impossible. Main reason is an almost 30 Ma gap in land tetrapode layers from Upper Triassic to Basal (?) Jurassic. The gap is not full but serious.
The first family tree is for final therapsids-basal mammals, Fig. 105 of Ref. 4. In early Middle Triassic completely herbivore tritylodonts detach from the main lineage. Latter at the end of Middle Triassic forks to tritheledonts and mammals. Comes the gap, and when good Jurassic layers appear, only tritylodonts and mammals can be recognised; tritylodonts die out with Middle Jurassic Stereognathus. So Jurassic mammals are descendants of either Upper Triassic mammals, or tritheledonts or both. (Consult again with the Table of appearance of mammalian characters in Chapter 2.)
A quite recent cladogram until "first mammals" is in :
Table 6: A list of groups branching out from the line leading to Morganucodon ("reptilian" side). Morganucodon is often considered "first mammal".
Morganucodon is first mammal, and then tritheledonts is the sister group of all mammals.
The second cladogram is compiled from Kemp + various recent articles. From the line to us the following main lineages detach, in temporal order:
_________end of Pantotheria----------
Table 7: As Table 6, but on the line "end-therapsids"-Eutherians.
See e.g. , . According to  Zhangheotherium (very Late Jurassic) is basal therian. This fossil supports sister-taxon relationship between multituberculates and therians. Zhangheotherium's forelimb is not too sagittal (compare to Monotremata!). For audiatory system. still has the primitive finger-like promontorium of non-therians, possibly with straight cochlea. The coiled therian cochlea seems later evolution, parallel with monotremes (partially coiled).
Finally, the inner structure of subclass Theria needs some attention. Pantotheria is sister group to the common ancestor of Metatheria and Eutheria. The next branching gives Metatheria and Eutheria as sister taxa. Metatheria results in Holoclemensia, Deltatheria and Marsupialia; maybe the first asundered first, and the latter two are sister taxa . Among Eutheria (defined by tooth structure + dentition) there are some early groups which were not placentals (e.g. Prokennalestes & Ukhaatherium) and then appears the placenta .
1) Marsupials are not synonyms of having marsupium. Marsupial bone was found in some tritylodonts, monotremes, multituberculates, deltatherids and eutheres (Ukhaatherium).
2) Eutheria and Placentalia are not synonyms either.
3) Marsupial Perameloidea acquired a kind of placenta independently of placentals.
4) According to Bonaparte Xenarthra (<Edentata) have multituberculate affinity (see next Chapter).
6. MAY EDENTATA BE NOT EUTHERIA?
If not all eutherians were placental, then it is not impossible that Xenarthra are not placentals but also acquired placenta independently, in which case there is no evidence that they would be eutheres. Namely Eutheria are defined through teeth, and Xenarthra are either toothless or if not, they have very special teeth.
Edentata differ from all living and most extinct mammals in a number of fundamental characters (Chapters 1 & 2). We have an alternative; both solutions problematic:
A) Edentata peculiarities are preserved from an old independent ancestor. But then Edentata lineage is independent from before Procynosuchus (Chapter 2), contradicting molecular clock.
B) Edentata peculiarities are synapomorphies compared to all other mammals. However for this a long independent evolution is needed, which is then almost Case A.
Now comes a brief summary of J. Bonaparte's suspicions about the origin of Edentata (or Xenarthra):
See first the original description of a Palaeocene animal which seems to be ideal precursor of some Xenarthra (they start from Late Palaeocene): Sudamerica . Then Bonaparte puts them into context: he finds Gondwanatherium from Campanian, which seems similar to Sudamerica. His tentative classification in 1986 :
Subclass Theria Parker and Haswell 1897
?Infraclass Paratheria Thomas 1887
He states that "Molars lack any sign of a tribosphenid cusp pattern", and guesses origin from eupanthoteres.
In 1990 Bonaparte's opinion is slightly different . It is simpler to cite it from an 1993 paper : "Most workers regard edentates as the most primitive known placental mammals...However, assuming that Sudamerica and Gondwanatherium were involved in the ancestry of Edentata, Bonaparte ... postulated <that is > that...edentates may be relicts of an endemic radiation which evolved in isolation prior to the differentiation of placentals from some primitive therian stock." However "Ferugliotherium is probably most closely related to an undescribed plagiaulacoid from the Upper Jurassic Morrison Formation". So Ref. 13 guesses that Superfamily Gondwanatherioidea belongs to Edentata, and so Edentata are very derived Multituberculata:
The debate is still continuing. But only that is important that such a view is possible. Reconstructions obviously differ if strange and still extant Edentata are not eutheres but multituberculates.
The next chapter may give a hint for the Edentata problem.
7. HOX GENE EXPRESSIONS AND MAMMAL EVOLUTION
Edentata (or at least Xenarthra) are peculiar also from amino acid sequences: they share a unique mutation . A 3-amino acid deletion in alpha-crystallin in eye lens is found in 2 sloths, 3 anteaters (Pilosa) and 2 armadillos (Cingulata). This deletion was not found otherwise, in: 55 sp. of 16 eutherian orders, 2 sp. of marsupials and 34 sp. of non-mammal vertebrates. So Xenarthra are probably monophyletic (and here is a strictly xenarthran apomorphy).
Then let us look at the Hox genes. Hox(a,b,c,d)-X genes are expressed in a linear sequence; the anterior border is sharp.
Change bw. therapsid/mammal
Mandible, middle ear
Hoxa-2 error: "reptilian"
Terminal skull bones
Error: wrong atlas
Anterior cervic vertebrae
Posterior cervic vertebrae
Not 7*. Edentata!?
Error: rib on 1st lumbar#
Error: reptilian uterus+
Error changes penis$
* Errors can transform cervic vertebraes into thoracic or vice versa.
# Thoracic/lumbar border is constant from higher therapsids (diaphragm!).
$ Probably did not change on therapsid/mammal border.
Table 8: Hoxy-X gene expressions and their possible effects on "mammal characters". X from 1 to 13, y=a, b, c, d. For details see , ,  & .
Then there may be a way to decide if Edentata peculiarities are synapomorphisms or not if Edentata Hox genes will be mapped.
8. CONCLUSIONS PRO TEMPORE
Parallel evolution (maybe also convergence) seems to have been a commonplace among therapsids in Triassic and Early Jurassic. All "distinctive mammalian" characters suggested until now turns to be continuously evolving or appearing multiply. Maybe the evolution of auditory organ will be simpler . Indeed it is closely correlated with jaw/skull articulation which is one of the classical distinctions (Chapter 2). Both promontorium and cochlea were growing in the sequence Adelobasileus, Sinoconodon, Morganuconodon, multituberculates & Haldanodon. But: tritheledonts were not investigated.
So the "first mammal" is now a stronly definition-dependent notion, and we still cannot be sure is one or more lineages crossed the "reptile-mammal border". For example, is it not possible that strictly herbivore multituberculates originate from strictly herbivore tritylodonts? Among synapsids new lineages started always with carnivores  and we do not know carnivore ancestors for multituberculates, tritylodont Oligokypus had more parasagittal limbs that much later Zhangheotherium or recent Ornithorhynchus, and tritylodonts had maternal care, because they had marsupial bone .
But even if we cannot tell who was the first mammal, the pattern of synapsid/diapsid competition is now clear. But let us first close the question of P/T "impact".
From the more pronounced extinction one would expect a crater significantly greater than Chicxulub. True, the event is older. But craters of a big Devonian impact in Scandinavia are still recognisable.
On the C/T boundary "iridium-rich" layers can be found at a number of spots world-wide. Iridium is suppressed in the terrestrial crust, so a more or less cosmic abundance (say, as in a C1 chondrite) albeit small in absolute numbers is too high in terrestrial context. Such arguments established the statement that at C/T extraterrestrial matter reached Earth, in significant quantity . Such a phenomenon is not observed at P/T. Superanoxia is observed . The absence of the normal terrestrial atmospheric O2 might be explained by the arrival of an agressive positive (so chemically metallic) matter, but such is not seen. On the other hand we do know the mass activity of Siberian volcanoes at end-Permian , ; this may have polluted the atmosphere. Now, we do not know the reason of this activity. At C/T boundary the (definitely smaller) synchronised Dekkan volcanic activity is attributed to the impact. At this moment it is better to say that the extraterrestrial matter is not found at P/T.
Then remains the third signal: spherules. The impact always produces spherules, and some of them are quite persistent. From South Africa, in the Fig Tree Formation 3.4 Ga old spherules are still recognisable . Generally only 1-3% of the spherules is extraterrestrial matter, the majority is target material, so it is quite possible that the extraterrestrial component is hidden. On the other hand, not all spherules come from impact.
Again using C/T as analogy, on C/T boundary spherules are known . Now, what is the situation for P/T spherules?
P/T spherules do exist. There are Miono-type spherules from Japan and China, and at least in one case the spherules dramatically coincided with the P/T boundary , . In the Sasayama rock wall the P/T layer is available and in 1997 rock were collected to get spherules but that work is still unpublished. A spherule is Miono-type if shape and size obey narrow constraints and from the Bükk Mts., Hungary, a small number of Miono-type spherules are reported. The chemical composition of Miono-type spherules is still not too intensively published; from Japan some 90% Fe is reported as metallic atoms. For the Hungarian Miono spherules still only photographs seem to exist. There is one more report, from Kashmir. Azmi reported one spherule from the boundary  which, however, did not seem Miono-type. The result was so fresh in autumn of 1999 that it was not in the written text .
Now we can get at least constraints to the catastrophic P/T event. Data do not exclude an impact but do not confirm either. If Miono-type spherules are mainly FeO, then they cannot come from an impact because there is no asteroid from mainly FeO. Metallic Fe, and then oxidation, causing anoxy, is not impossible, but then all spherules would contain impactor material, and no target material would be found; a situation never met in impacts. Then probably (?) the event was not an impact but a "collision" with some galactic cloud. But details are still mainly unknown.
Marine sediments show the anoxy. I think that the size reduction at therapsids at P/T and later in Triassic do show the decrease of oxygen level in the atmosphere, but not a superanoxy: the animals with the most intensive metabolism of the age, therapsids, did not die out, albeit reduced size. It seems that on land the extinction was far from 97%; and be careful. The "event" may have lasted 10 million years . After 10 million years the majority of the original species are substituted by new ones even without catastrophes.
No doubt, there was a catastrophe near P/T, but also, it seems that we do not yet know the details to sufficient details. However the pattern is clear. Mid-Permian cooling at temperate zones preferred endothermic animals with efficient feeding, hair, good respiration (e.g. diaphragm) and so on. When this started, our synapsid ancestors were just the lords of ancient world: the largest land animals. They met the challenge and answered it properly.
Only when they were just adapting, the temperature went back, and oxygen level down, for any reason. (The transient cooling may have been caused by the cyclic operation of Sun, mentioned in ; I repesat that I do not know the reason of decrease of oxygen level.) So Nature goated our ancestors into a cul-de-sac and then betrayed them. The poikilotherm lazy diapsids returned from the Equator (true reptilian metabolism can tolerate lower oxygen level) and when our ancestors were shrinking to improve surface/volume ratio, they could expand. In uppermost Triassic "first mammal" (anything it was) and first dinosaur appeared practically simultaneously. But that first mammal was of mouse size.
Ice went away too soon. Mesozoic had not become the Age of Mammals. It was the Age of Dinosaurs and our ancestors were tiny mice amongst their legs.
In Cretaceous oxygen level was high, so mammals could grow; but a too big mammal still meant a dwarf from dinosaurian viewpoint: not an enemy but already a possible prey. Only multituberculates could reach cat-size and we do not know how they reached a modus vivendi with the big diapsids because now they are extinct so cannot tell us (if Xenarthra are not multituberculate offshot). Anyway, with the moderate size expansion multituberculates put themselves out of the inter-synapsid competition (so effective in Late Permian and Triassic, leading to parallel evolution, coevolution &c. of therapsids and mammals). Multituberculates were the biggest synapsids but herbivores. They did not attack smaller mammals and mammalian carnivores were much smaller.
Then came the Dinosaur Killer. Earth was given back to her rightfull owners the synapsids. They expanded in size and occupied niches left empty by dying big diapsids. And what is interesting: multituberculates, the relative most successful synapsids of Mesozoic, were unable to accomodate themselves to a new, synapsid world.
For today this competition is practically over. Diapsids are represented by tuatara who lives only on islands off-shore New Zealand; by snakes & lizards who may be enemies but in general not competitors; and by birds, generally small and in the skies. On land we dominate and it seems that there will not be a third round.
But mammal intelligence (e.g. this lecture) was delayed with dozens of million years during the 120 million years of Dark Ages of mesozoic diapsids.
APPENDIX: A PROBLEM WITH DNA DISTANCES
DNA sequences should be excellent for discover the natural system of living mammals. However it seems that there are problems. One example is aardvark . The side product is a strange eutherian tree. It starts nicely, with a trifurcation (Monotremata, Metatheria, Eutheria). (Compare with Ref. 1, which predates ). But on the tree we (Primates) are sister group of all other Eutheria, including Xenarthra (bats are not on the tree).
Now, J. D. Pettigrew in a Chiroptera lecture  mentions the A+T bias falsifying some DNA distances. Namely, if DNA sequences are not explicitly dependent on environment or lifestyle then it is difficult to explain that (A+T) percentage is monotonously increasing with activity. Namely
average or cca.
Table 9: A+T percentage in the genetic matter. If natural selection did not react back to DNA & RNA, the expected number would be cca. 50. By redigitalising Slide 47 of Pettigrew . It seems as if more active lifestyle were to prefer higher A+T. Satisfying explanations are not available.
The effect is difficult to interpret; but it may indicate possible distortions in genetic distances.
This text is the somewhat edited form of the lecture at the Geonomy Scientific Committee of the Hungarian Academy of Sciences by the president of the Matter Evolution Subcommittee of the said Committee. The purpose of manufacturing this written version was first documentation, second to collect a reference list.
I would like to announce that a loose unofficial research group (Mammalia Group) came into hazy existence, with some connection to the Matter Evolution Subcommittee. Members (affiliations) are: Sz. Bérczi (Dept. General Physics, ELTE, Budapest), Agnes Holba (CRIP RMKI), B. Lukács (CRIP RMKI), I. Molnár (Dept. of Genetics, ELTE, Budapest) and Éva Papp (AGSO, Canberra). The group believes that some steps in mammal evolution still need clarification. I acknowledge the members' help in preparing this lecture.
Also I have to declare that I was a member of IGCP 384 Spherules and Global Events, which gave me some insight into P/T spherules. However my connection to the Hungarian leadership of that program does not exist since the beginning of 1999, although they continued to the Japanese side. All discussions here about Permian and P/T extraterrestrial influences are part of the activity of OTKA T/026660 "Thermodynamics os Asteroids and Meteorites, whose members are Sz. Bérczi, Agnes Holba, B. Lukács and K. Martinás. I acknowledge their positive influence too.
Illuminating discussions with Sh. Miono & Y. Miura are acknowledged.
 Sz. Bérczi, Agnes Holba, B. Lukács & Éva Papp, KFKI-1998-07, p. 20
 S. Kumar & S. B. Hedges, Nature 392, 917 (1998)
 Sz. Bérczi & B. Lukács, Acta Climat. 31A, 37 (1998)
 T. S. Kemp, Mammal-like Reptiles and the Origin of Mammals, Academic Press, London, 1982
 Z. Luo & A. W. Crompton: J. Verteb. Paleont. 14, 341 (1994
 Q. Ji, Z. Luo & Sh. Ji: Nature 398, 326 (1999)
 Y. Hu, Y. Wang, Z. Luo & Ch. Li: Nature 390, 137 (1997).
 G. W. Rougier, J. R. Wibble & M. J. Novacek: Nature 396, 459 (1998)
 M. J. Novacek & al.: Nature 389, 483 (1997)
 G. J. Scillato-Yané & R. Pascual: Ameghiniana 21, 173 (1985)
 J. F. Bonaparte: J. Vert. Paleont. 6, 264 (1986)
 J. F. Bonaparte: Natl. Geogr. Res. 6, 63 (1990)
 D. W. Krause & J. F. Bonaparte: Proc. Natl. Acad. Sci. USA 90, 9379 (1993)
 Marjon A. M. van Dijk, E. Paradis, F. Catzeflis & W. W. De Jong: Syst. Biol. 48, 94 (1999)
 J. Zákány & al.: Proc. Natl. Acad. Sci. 94, 13695 (1997)
 Ann C. Burke & al.: Development 121, 333 (1995)
 Nina M. Brooke & al., Nature 392, 920 (1998)
 M. Mark, F., M. Rijli & P. Chambon: Pediatric Research 42, 559 (1997)
 Z. Luo, A. W. Crompton & S. G. Lucas: J. Verteb. Paleont. 15, 113 (1995)
 L.W. Alvarez & R. A. F. Dence: Science 208, 1095 (1979)
 Y. Isozaki: Pangea: Global Environment and Resources, Canadian Society of Petroleum Geologists, Memoir 17, 805 (1994)
 P. J. Conaghan, S. E. Shaw & J. J. Veverka: Pangea: Global Environment and Resources, Canadian Society of Petroleum Geologists, Memoir 17, 785 (1994)
 D. R. Lowe & al.: Science 245, 959 (1989)
 Y. Miura, T. Kato & M. Imai: Celestial Mech. Dyn. Astr. 54, 249 (1992)
 Sh. Miono & al.: Nucl. Instr. & Meth. B75, 435 (1993)
 Miono Sh.: Impact and Extraterrestrial Spherules (Japan Contibution to the IGCP) 1998, p. 15
 R. J. Azmi: Lecture at PIECE'99, Yamaguchi, Sept. 27, 1999
 R. J. Azmi: PIECE'99, ed. Y. Miura, Yamaguchi, 1999, p. 4
 B. Géczy: Annales Univ. Sci. Budapestiensis R. Eötvös, Sect. Geophys & Meteorol. XII, 13 (1996)
 U. Arnason, Anette Gullberg & A. Janke: Proc. Roy. Soc. London B266, 339 (1999)
 J. D. Pettigrew on Internet, title "404 Not Found" (sic!), http://www.uk.edu.au/nuq/jack/consensus/htm
My HomePage, with some other studies, if you are curious.