1923 CHRONOLOGY 47 



occur in one bed (see Pis. CCCLIV b and CCCLXXXIII b) ; but it 

 gives P. cf. devillei in a higher stratum than Am. pectinatus, and there 

 is reason to suppose that Paravirgatites is from a still higher horizon 

 (see PL CCCVIII b) — at any rate, it has another distinct matrix. 

 So there is this result : Oxfordshire for four forms shows two matrices, 

 Swindon, for the same four forms, three matrices ; but Swindon parts 

 the three which Oxfordshire puts together, and Oxfordshire parts the 

 two which Swindon unites — thus proving the four forms to be in sequence, 

 so that four hemeral terms are required — three on the evidence of 

 Swindon, two on the evidence of Oxfordshire, minus the one in common, 

 makes four. 



The argument from dissimilar matrices — ^more dissimilar, aniso- 

 petrous — may be presented. Two authors describe sequent beds at 

 locahties A, B, some distance apart. One author describes the lower 

 deposit, .-1, which is argillaceous ; the other author describes the higher 

 horizon, B, which is calcareous. Examination of the figured specimens 

 shows that in the main the two Ammonite faunas are distinct, but that 

 some 25 per cent, are common — the one author claiming them as con- 

 stituents of the clay, the other of the limestone. To say that the common 

 fauna passed up from one deposit to the other is incorrect. The true 

 answer is that three faunas have been dealt with, a, the earUest, b, the 

 common fauna, c, the latest. Thus there are three hemerae, a, b, c, 

 and during hemera b clay deposits prevailed at locahty A, but calcareous 

 deposits had begun at locahty B. Therefore the constituents of the 

 b fauna should be found in the highest clay beds of locahty A, and in 

 the lowest hmestone beds of locality B. It may, further, be predicted 

 that the fauna b will not be found in the calcareous beds of A» nor in 

 the argillaceous beds of B. 



This argument from dissimilar matrices can be employed to predict 

 the hemeral sequence of a given fauna when hemeral or zonal analyses 

 have not been carried far enough. Some years ago, because certain 

 Ammonite species were found in different parts of the south-west of 

 England in clay, sand and limestone, which are sequent throughout the 

 region, it was said that these Ammonites passed up through three 

 formations. Detailed investigation showed that such was not the case — 

 that there were always the same sequences of species, that clay, sand 

 and hmestone were being deposited simultaneously at different locahties, 

 and that there was no case of passing up in any one locality. 



It is from considerations such as these, where direct evidence of 

 superposition was lacking, that the hemeral tables have been constructed. 

 But that the Tables are free from mistakes is too much to expect. 



The illustrations of Ammonites which have been given in the four 

 volumes of this work are intended not only for the use of the speciahst, 

 but for the assistance of any student of Mollusca seeking to identify 

 the specimens in his collection. For this reason there has been given in 

 each case, at the top of the legend footing each plate, the name which 

 the species has borne in literature, or, failing that, the name which it 

 has received in pubhc or private collections, or, faihng that, the name 

 which has or might have been applied to it by the field-geologist. 

 Therefore, if the student is aware of the name which has hitherto been 

 appUed, even in a general way, to the specimen which he is seeking to 

 determine, he can look up that name in the index, and will find references 

 to the plates which have been given of the species bearing that name. 

 Thus, instead of aimlessly turning over plates, only to become more and 

 more bewildered by a seemingly endless array of forms, the student can, 



