24 



SCIENCR 



[Vol. XXI. No. 519 



stated that y^'^^ part of antimony will convert the best select cop- 

 per into the worst conceivable. Another instance occurs in the 

 •case of iron. By the addition of ,% per cent of carbon steel is 

 produced of such a kind as would make an excellent bridge, or 

 boiler plate, but if fashioned into a weapon would be absolutely 

 untrustworthy If, on the other hand, /j per cent of carbon were 

 introduced, a material is obtained from which a good razor might 

 be made, but it would be useless for a rail or the construction of 

 a bridge. A Irace of manganese in steel renders it impossible to 

 make a magnet out of such a specimen. It also prevents the 

 hardening of such steel by rapid cooling after heating to redness. 



The metal, however, which shows the aiost remarkable change 

 in its physical properties when contaminated with next-to-noth- 

 ing of a foreign substance is gold. The addition of j^ per cent of 

 bismuth would render a specimen of gold useless for coinage pur- 

 poses, as it would crumble to powder under the pi'essure of the 

 die. Lead acts in a similar way. One part of lead added to two 

 thousand parts of gold reduces its tenacity from 18 tons per square 

 inch to only 5 ton's. A bar of such gold can be readily broken by 

 a tap from a hammer. The color of the gold is changed from 

 yellow to orange brown. Such a remarkable change in the ap- 

 pearance and properties of gold on the addition of small quanti- 

 ties of other substances was known in the seventh century and 

 helped to confirm the belief of the alchemists that they had only 

 to find some substance which would alter the properties and ap- 

 pearance of any given metal so that it would change into and ac- 

 ■quire the properties of gold. Hence the search for the philoso- 

 pher's stone. 



This paper might be indefinitely extended, but enough has prob- 

 ably been said to show that even in chemistry the day of small 

 things is not to be despised, and that a thorough investigation of 

 some of the commonest and best-known chemical changes would 

 ■doubtless bring to light many facts at present overlooked, and 

 ■would tend to a better understanding of the workings of nature. 



• BREAD-FRUIT TREES IN NORTH AMERICA. 



BY F. H. KNOWLTON, V. S. NATIONAL MUSEUM, WASHINGTON, D.O. 



The living species of the genus Artocarpus are exclusively Old 

 World, being confined in their distribution to tropical Asia and 

 the Malay Archipelago. About forty species have been described, 

 ■of which number two or three are now widely cultivated through- 

 out the tropics, the most important of these being A. incisa, the 

 true bread-fruit tree. They are small or medium-sized trees with 

 a milky juice, and large, leathery, entire, or pinnately lobed, or 

 rarely pinnately compound leaves. The flowers are monoecious 

 with the staminate ones borne in long club-shaped spikes, and the 

 pistillate in rounded heads. The female flowers soon grow to- 

 gether and form one large, fleshy mass, or the so-called bread- 

 fruit. When mature, the fruit is marked on the exterior with 

 hexagonal knobs, and in the interior consists of a whitish pulp, 

 having the consistence of new bread, whence its name. 



Although not at present an element in the flora of the New 

 World, there is now abundant evidence to show that the genus 

 Artocarpus was, during late Cretaceous and earlier Tertiary times, 

 an inhabitant of North America. The best known species, called 

 Artocarpus lessigiana (Lx.), was discovered in 1874 in the Lower 

 Laramie on Coal Creek, in Boulder County, Colorado. It was 

 ■first described by the late Professor Leo Lesquereux, under the 

 name of Myrica 1 lessigiana, on the supposition that it was a 

 gigantic representative of the genus Myrica. Specimens, now 

 known to represent the upper portions of large leaves, were later 

 obtained from the andisitic deposits forming the recently differ- 

 entiated Denver formation of South Table Mountain, near Golden, 

 Colorado. These leaves were called Aralia pungens by Professor 

 Lesquereux, who naturally confounded the imperfect examples 

 at his disposal with well known fossil forms of this genus, which 

 they much resemble. Since that time several additional speci- 

 mens have been obtained, which not only prove that Blyrica ? 

 lessigiana and Aralia pungens are identical, but also that they 

 :should be referred to Artocarpus. 



The leaves of Artocarpus lessigiana were very large, measuring 

 30 centimeters in length and 18 or 20 centimeters in width. They 

 are thick, probably coriacious in texture, broadly oblong in gen- 

 eral outline, and deeply, pinnately 4-6-lobed. The lobes are ob- 

 long, lanceolate, taper-pointed, and separated at the base by 

 broad, rounded sinuses, the lobation being most extensive at the 

 base of the leaf, where the sinus almost reaches the midrib, and 

 the two lower lobes are connected by a narrow ring only. The 

 nervation of the leaf is very strong, and precisely like that of the 

 living A. incisa, which differs from the fossil in having the deep- 

 est lobation in the upper part of the leaf. 



Closely allied to this species, and possibly identical with it, is 

 what I propose to call Artocarpus californica, which is founded 

 upon specimens obtained by Dr. Cooper Curtice, then of the U. S. 

 Geological Survey, from the auriferous gravels at Independence 

 Hill, Placer County, California. This species differs from the 

 former by its smaller size, thinner texture, and shorter, more 

 acute, lobes. It is not suflSciently well preserved to show the 

 finer nervation, but, as far as can be made out, it is very similar 

 to A. lessigiana, and additional material may show them to be 

 the same. 



Specimens, probably belonging to this species (A. californica), 

 were obtained some years ago from the John Day Valley in Ore- 

 gon, the age of which is either Upper Miocene or Lower Pliocene. 

 They were identified by Professor Lesquereux both with his 

 Myrica? lessigiana and Aralia pungens; but, as they are some- 

 what fragmentary, it is not possible to be positive as to their cor- 

 rect determination. 



The most northern point at which the genus Artocarpus has 

 been found fossil is northern Greenland, in latitude 70'^. Dr. A. 

 S. Nathorst obtained a large leaf, which he named A. dicksoni, 

 in the Cenomanian near Waigatt. This species is also closely re- 

 lated to A. incisa, and was associated with a fruit which is un- 

 questionably that of a bread-fruit tree. Nathorst, who was the 

 first to point out the true relationship of Lesquereux's Myrica ? 

 lessigiana and Aralia pungens, suggests the possibility of their 

 being the descendants of the Greenland species, which may 

 have deen dispersed over the North American continent by the 

 ice-sheet. The material at present available is hardly sufficient 

 to establish unquestioned relationship between them, for the ner- 

 vation of A. dicksoni is not to be made out, but, as all are un- 

 doubtedly related to the living bread-fruit (A. incisa), they may 

 be more closely related among themselves than now seems ap- 

 parent. 



From the above account, it appears that the bread-fruit trees 

 existed in North America as far north (in Oregon) as 46°, and as 

 late as early Pliocene or late Miocene time. The reason for their 

 complete disappearance from the American flora, and that within 

 such a comparatively short space of time, is diflScult to supply. 

 If they had been pushed southward, and now inhabited the tropics, 

 it would be readily explainable, and quite in accord with other 

 well-known instances, but they have totally disappeared from 

 the New World, notwithstanding the fact that they grow when 

 transplanted as freely in tropical America as in their native coun- 

 try. It is probable that the advance of the refrigeration was so 

 rapid that they were unable to escape in the New World, and 

 perished to the last one, while in the Old World some avenue per- 

 mitted their perpetuation. The genus Eucalyptus is another ex- 

 ample of the same condition. During Cretaceous and Tertiary 

 times it was an inhabitant of North America and Greenland, but 

 is now entirely confined to Australia. 



The deductions to be drawn, as to the climate that prevailed at 

 the time when these trees existed in North America, are to be 

 made with caution. The fact that all the living species of a 

 genus are tropical does not necessarily prove that it has always 

 been so. Again, a genus that is essentially tropical may have 

 species extending into sub-tropical or even temperate regions. 

 The genus Dicksonia is a marked example of this kind. It is 

 principally an inhabitant of tropical America and Polynesia, but 

 one species reaches as far north as Canada, and several are scat- 

 tered throughout the southern part of the temperate zone. 



Taken by itself, Artocarpus would indicate a tropical climate, 

 but the plants with which it is associated have also great weight 



