Boyer diss 2009 1046..
Boyer diss 2009 1046.. Boyer diss 2009 1046..
zero or less were excluded from geometric mean calculations (e.g., the distance that the greater trochanter of the femur projects proximal to the femoral head could easily be positive, zero or negative in a taxon of any absolute size). Particular variable sets used to create geometric means are cited in relevant table legends below. The resulting geometric mean ratios were then natural-log transformed. Whether it is necessary to use logarithms of ratios is debated (e.g., Jungers et al., 1995). I chose to do so here based on the fact that it frequently improves normality of distributions of ratios (Sokal and Rohlf, 1997). Furthermore my own experimentation with different data formats lead me to use logs rather than raw ratios. For instance, a version of the analysis of humeral variables performed on un-logged ratios captured less variance, and captured it less evenly in the first two coordinates, than the logged version of the same data (see additional discussion in Chapter 5). Body mass estimation.—Body mass is estimated for P. cookei and other plesiadapids using six limb bone lengths and six diameters. Two different sets of regressions are used: Gingerich (1990) generated regressions from data of Alexander et al. (1979). This dataset samples a phylogenetically diverse array from shrews to elephants. An unpublished dataset of primates from the UMMZ collection is used to generate a parallel set of regressions. The results of these two sets of regressions are presented and compared. Intrinsic hand proportions.—A reconsideration of digit proportions of P. cookei using methods and the comparative data set of Kirk et al. (2008) is undertaken. 280
Organization of results Each bone is discussed in its own Results section. Each of these sections has as many as three subsections: description, function, and comparisons. The first of these subsections is always the description. Descriptions include minimal comparative and soft anatomical references. Subsections assessing functional features of each bone and comparing P. cookei to other extant and fossil taxa (mainly other plesiadapids) often follow each description. Following the results sections focusing on specific bones, results of comparative analyses that integrate the morphology and measurements from many regions of the skeletons are presented (i.e., those analyses that cannot logically be included under the heading of a particular bone or set of bones). 281
- Page 257 and 258: Figure 3.6. Cranium of Plesiadapis
- Page 259 and 260: Figure 3.8. Fragment from right nuc
- Page 261 and 262: Figure 3.9. Right promontorium of P
- Page 263 and 264: Figure 3.10. Cranium of Plesiadapis
- Page 265 and 266: Figure 3.12. Right dentary of Plesi
- Page 267 and 268: Figure 3.14. A, Plot of relief inde
- Page 269 and 270: CHAPTER 4: THE FIRST KNOWN SKELETON
- Page 271 and 272: among plesiadapiforms (e.g., Szalay
- Page 273 and 274: Institutional and collections abbre
- Page 275 and 276: CaL - capitulum (of humerus) antero
- Page 277 and 278: HSV - head shape variable = ln(DEW/
- Page 279 and 280: MSD - mid-shaft dorsoventral or ant
- Page 281 and 282: Ry - ray (as in “digit ray”) S-
- Page 283 and 284: History of descriptive study of the
- Page 285 and 286: illustrations of this material, exc
- Page 287 and 288: astragalus and calcaneum was highly
- Page 289 and 290: discussion of the femur indicates t
- Page 291 and 292: supinator crests. He also noted tha
- Page 293 and 294: that it may not even be an archonta
- Page 295 and 296: unstudied material. Specifically, h
- Page 297 and 298: 5321), some metapodials (MNHN R 529
- Page 299 and 300: Gingerich and Gunnell (1992) publis
- Page 301 and 302: prehensility they provide, is an in
- Page 303 and 304: euarchontans (Fig. 1.1). Their anal
- Page 305 and 306: for comparison. These include isola
- Page 307: plesiadapid samples have the same m
- Page 311 and 312: Bloch and Boyer (2002) and N. inter
- Page 313 and 314: clavicle reflects some basic aspect
- Page 315 and 316: Humerus Description.—The right an
- Page 317 and 318: epicondyle actually projects somewh
- Page 319 and 320: cookei is absolutely longer than an
- Page 321 and 322: tuberosity. This crest probably del
- Page 323 and 324: olecranon process to estimate its t
- Page 325 and 326: distinct, convex distal radial face
- Page 327 and 328: of the midcarpal joint), and its pr
- Page 329 and 330: (there is no evidence for more than
- Page 331 and 332: matches the opposing facet on the t
- Page 333 and 334: mobility at the trapezoid-trapezium
- Page 335 and 336: Function.—The three proximal carp
- Page 337 and 338: the bone presently being described:
- Page 339 and 340: the “set 2” MC II is a larger,
- Page 341 and 342: differs from MC II and III in havin
- Page 343 and 344: even more pronounced. The distal en
- Page 345 and 346: etween the distal carpals and the
- Page 347 and 348: have stouter shaft diameters for th
- Page 349 and 350: difference makes them more like kno
- Page 351 and 352: antipronograde clinging postures, o
- Page 353 and 354: foramina, and faces slightly proxim
- Page 355 and 356: spine at the superior tip of the il
- Page 357 and 358: the thigh (Gambaryan, 1974). The ha
zero or less were excluded from geometric mean calculations (e.g., the distance that the<br />
greater trochanter of the femur projects proximal to the femoral head could easily be<br />
positive, zero or negative in a taxon of any absolute size). Particular variable sets used to<br />
create geometric means are cited in relevant table legends below. The resulting<br />
geometric mean ratios were then natural-log transformed. Whether it is necessary to use<br />
logarithms of ratios is debated (e.g., Jungers et al., 1995). I chose to do so here based on<br />
the fact that it frequently improves normality of distributions of ratios (Sokal and Rohlf,<br />
1997). Furthermore my own experimentation with different data formats lead me to use<br />
logs rather than raw ratios. For instance, a version of the analysis of humeral variables<br />
performed on un-logged ratios captured less variance, and captured it less evenly in the<br />
first two coordinates, than the logged version of the same data (see additional discussion<br />
in Chapter 5).<br />
Body mass estimation.—Body mass is estimated for P. cookei and other<br />
plesiadapids using six limb bone lengths and six diameters. Two different sets of<br />
regressions are used: Gingerich (1990) generated regressions from data of Alexander et<br />
al. (1979). This dataset samples a phylogenetically diverse array from shrews to<br />
elephants. An unpublished dataset of primates from the UMMZ collection is used to<br />
generate a parallel set of regressions. The results of these two sets of regressions are<br />
presented and compared.<br />
Intrinsic hand proportions.—A reconsideration of digit proportions of P. cookei<br />
using methods and the comparative data set of Kirk et al. (2008) is undertaken.<br />
280