MEMOlliS OF THE NxVTIONAL ACADEMY OF SCIENCES. 17 



widespread cliansc in tlie oiiviroiiineiit, iiivolvinjj' a dryiii!;' ii)> of the, soil, much of it alkaline, the 

 direct iutluence on plant life must have been profound, as regards their protective defenses, and 

 after spines began to develop one can well understand how their shapes should have been regulated 

 for each species and preserved by the set of niinor factors wliich pass current under the term 

 "natural selectio'n."' 



Animals may also, in some cases, have developed spines in response to a change of environ- 

 ment. If we ghince over the eiinclis of paleontological history we shall see that at certain periods 

 trilobites, brachiopods, ammonites, and perhaps other groups showed a tendency to become tuber- 

 culated, spiny, or otherwise excessively ornamented. These periods must have been characterized 

 by great geological changes, both of the relative distribution of land and water and perhaps of 

 climate and soil. Among the brachiopods, more spiny species .occur in the Carboniferous period 

 than in the earlier Paleozoic times.' Among the trilobites, although in Paradoxides and in other 

 genera the gensB and sides of the segments are often greatly elongated, we only find forms with 

 long dorsal spines at the close of the Silurian and during the Devonian.'^ There are no such spiny 

 forms of ammonites as in the uncoiled Cretaceous Crioceras,^ etc. 



These types, as is well known, had their period of rise, culmination, and decline, or extinc- 

 tion, and the more spiny, highly ornamented, abnormal, bizarre forms appeared at or about the 

 time when the vitality of the type was apparently declining. Geddes claims that the spines of 

 plants are a proof of ebbing vitality. Whether or not this was the case with the types of animal 

 life referred to, whether the excess of ornamentation was due to excess or deficiency of fooil, it is 

 not improbable that the appearance of such highly or grotesquely ornamented forms as ce.tain 

 later brachiopods, trilobites, and ammonites was the result of a change in their environment during 

 a period when there were more widespread and profound changes in physical geography than had 

 perhaps pi'eviously occurred. 



If the tendency to the production of spines in past geological times was directly or indirectlj- 

 due to a change in the milieu, and if plants when subjected to new conditions, such as a transfer 

 to deserts, show a tendency to the growth of thorns, or if those which are constantly submerged 

 tend to throw out ascending aerial roots,* or if, like epiphytes, when growing in mid-air, they throw 

 out descending aerial roots, I have thought it not improbable that tubercles, humps, or spines may 

 have in the first place been developed in a few generations, as the result of some change in the 

 environment during the critical time attending or following the close of the Paleozoic or the early 

 part of the Mesozoic age, the time when dei'iduous trees and flowers probably began to appear. 



I have always regarded the Bombyces, or the superfamily of silkworm moths, as a very 

 ancient one, which has lost many forms by geological extinction. We thus account for the many 

 gaps between the genera. Both the larvai and the moths differ structurally far more than the 

 genera of Geometrids and of Noctuidte, and the number of species is less. The two latter families 

 probably arose from the great specialization of type in Tertiary times; while evidently the great 



' Although there are spiny brachiopods in the Silurian, they become more common in the Devonian (e. g., Atrypa 

 hi/strir, Chonetes scititla, C. coronata, C. muricala, Productella hirsiita, P. hiistricula, P. rarhpinn, an 1 Slrophacosia 

 yiroductoides), and are apparently more numerous in the Carboniferous formation (e. g., Proiiiirtiis Inngispinits, P. 

 nebrascensin, Chonelfs ornata, C. mcsoloba, C. variolala, C. salmanianit, C. seligcrus (also Devonian), C. fiaeheri, etc., 

 ProdurttUa newberriji, besiiles the Permian Productus horrida. 



-Besides Paradoxides, there are such forms as the Cambrian Hiidrocephalim carens, the Silurian Dalmtmia punctnia, 

 Cheirurns plnircxanthemiis, and Eunjcare breincauda, while the spiny species of Acidaspis seem U> be more abundant in 

 the Deronian than in the .Silurian strata, but those which bear dorsal spines, such as Deiphon forbesii and J iv/cs 

 annatua, are Devonian. 



^ Quite long spines occur in the Cretaceous species of Crioceras and Anci/loceras malheroiiinntim of Europe, but 

 none, so far as we .are aware, in earlier times. 



■•See N. S. Sh.aler: Notes on Taxodiiim dialichuin, Mem. M. C. Z., xvi, 1, 2, and W. P. Wilson: The production 

 •of aerating organs on the roots of swamp .and other pl.ants, Proc. Acad. Nat. Sci. Phil., April 2, 1889, quoted in 

 Garden and Forest, Jan. 1, 1890. Shaler conjectures that the function of the "knees" is in some way connected 

 with the aeration of the sap. Mr. Wilson shows that " besides the cypress, other plants which habitually grow with 

 roots covered with water (the water gum, Ni/ssa silvatioa, var. aqitatica, Avicennia nitida, and Pinna aerotina) develop 

 similar root processes; and what is still more suggestive. Mr. Wilson has induced plants of Indian corn to send roots 

 above the surface of the soil by keeping it continually saturated with water."' It is to be observed that the aerial 

 roots of the latter develop in a single generation. 



S. Mis. 50 2 



