Mabch 6, 1903.] 



SCIENCE. 



393 



energy to get their usual amount of grass. 

 Old Joe had even refused his grain for about 

 half the time. 



It should be remembered that our horses 

 had but a small amount of the grass. The 

 ranchmen told us that other travelers coming 

 into the coimtry had been obliged to camp 

 for a week while their horses slept off the 

 effect of a good feed of it, and while its ef- 

 fects usually lasted for a week or ten days, 

 it did no more serious damage than to leave 

 the animals thin from fasting. Stories were 

 told of horses being lost in the mountains 

 and found several days later in the bushes 

 near camp fast asleep. 



I have ofiered no real proof that this par- 

 ticular species of grass is what affected our 

 horses. They undoubtedly ate a dozen other 

 species of grass, as well as some other plants, 

 every day while we were in the mountains. 

 But after our experience I am inclined to 

 give credit to the uniform statements of the 

 ranchmen in regard to it. All agree on the 

 species, on its effects, and to the fact that 

 after one good dose of sleepy grass, horses 

 will never touch it again. This latter state- 

 ment has ample proof. Horses and cattle 

 are ranging in many of the valleys where it 

 grows in abundance, untouched and full of 

 ripe seed, while the other grasses are cropped 

 close all around it. I did not see horses or 

 cattle touch it except in the case of our own 

 animals and the team of another traveler 

 from the valley, all of which ate it eagerly. 

 They ate both the base leaves and the heads 

 that were full of ripe seeds. I shelled out 

 and ate a handful of the seeds, but without 

 noticeable effect. The ranchmen generally 

 agree that it is the leaves which produce the 

 sleepiness. 



I did not hear that cattle were affected by 

 it, but they certainly avoid it, as many were 

 grazing near where it stood untouched. 



While this experience was new to me, I 

 find that sleepy grass has long been known 

 to botanists as such, or technically as Stipa 

 vaseyi. Something has been known of its 

 effects on horses, but apparently its chemical 

 properties have not yet been determined. 

 Vernon Bailey. 



THE VERTEBRAL COLUMN OF BRONTOSAURUS. 



Although the genus Brontosaurus Marsh 

 has been known from the greater part of the 

 skeleton for more than twenty years, many 

 points of interest concerning its structure re- 

 main undetermined. The Field Columbian 

 Museum Expedition of 1900 was fortunate in 

 securing a large part of a skeleton of one of 

 these great reptiles in such a state of preserva- 

 tion that the bones of the torso and base of the 

 tail were scarcely disturbed from their relative 

 positions. This splendid specimen, which is 

 now almost ready for exhibition, makes it 

 possible to determine the vertebral formula of 

 the thoracic and anterior caudal regions, as 

 well as many other minor features. 



The specimen consists of eleven presacral 

 vertebra, five coalesced sacral, and twenty- 

 three caudal vertebrae, with pelvis, ribs and 

 chevrons almost intact. The eleventh pre- 

 sacral was exposed and partially broken away 

 when found. From that point backward the 

 thoracic, sacral and caudal vertebrae, as far 

 as caudal XIII., were lying in a close series, 

 with their centra nowhere displaced more than 

 two or three inches. Most of the ribs and 

 many of the chevrons were also found in 

 position. 



The specimen throughout agrees very close- 

 ly, both in size and in character, with Marsh's 

 type, Brontosaurus excelsus. However, it 

 shows that with regard to the thoracic region 

 his final restoration was considerably at fault. 

 In fact his first figure * shows the thorax 

 much more nearly correct. Counting the five 

 coalesced vertebrae as sacral, the thoracic series 

 in this specimen is made up of ten rib-bearing 

 vertebra. The eleventh, as before stated, has 

 been partially lost, but enough remains to 

 show that the transverse process is replaced by 

 a cervical rib. A noticeable reduction in size 

 of the rib facets on presacral X together with 

 the much-reduced neural spines on presacral 

 XI., bears out the conclusion that the latter is 

 the posterior cervical. We may, therefore, 

 conclude that the number of thoracic vertebrae 

 in this genus is ten instead of fourteen as 

 estimated by Marsh. 



The crest of the dorsal arch was evidently 



* Am. Jour. Soi., Vol. XXVI., pt. I. 



