This second version of our preprint has benefitted significantly from several very detailed, careful comments on the first version. We consider these least as substantial as most of the formal peer reviews our other works have received, and warmly thank everyone who contributed.
We have adopted all the suggestions of Emanual Tschopp (https://peerj.com/preprints/67/#feedback-67) and Mark Robinson (https://peerj.com/preprints/67/#feedback-69), with only trivial exceptions -- e.g. we continue to refer to Mannion et al. 2011 by that date even though, as Tschopp points out, the printed version of the paper came out in 2012, because we consider the initial (online) publication the important event.
Some of the comments of Andy Farke are more problematic, because they treat this paper in isolation whereas (unknown to Farke at that time) we have another paper on Barosaurus in preparation. Because of this, we feel that it is better to reserve a Systematic Palaeontology section and revised diagnosis for that paper, which introduces relevant new information. We do plan to do the work outlined by Farke -- but not in this paper.
We do not have, and cannot easily obtain, photographs of Cervical T, so cannot include a description and illustration as Farke requests. Lull (1919:14) says "The bone is only partly preserved, the fragment consisting of the posterior part of the centrum, bearing the left postzygapophysis and a portion of the right". It is unlikely that this element would yield meaningful information about Barosaurus cervical morphology.
We retain the illustration of the Giraffatitan cervical MB.R.2180:C5, not only or its significance to the present paper but also because there are no good published illustrations of any of the Giraffatitan material other than the line-drawings in Janensch's original papers.
Otherwise we have accepted Farke's suggestions.
The long comment from John Foster is also very helpful, but we have elected to link to it rather than incorporate his detailed argument into the manuscript, as it is tangential to the main thrust of the present work.
Mickey Mortimer makes several suggestions that would reault in a good and interesting paper on the phylogenetic distribution of neck-related morphologival characters in Sauropoda. But that paper is not this one: we want to keep this focussed on Barosaurus, while keeping open the option of subsequently working in the area that Mortimer recommends.
(Not directly relevant to the present paper, but Mortimer asserts: "including unique autapomorphies in analyses is largely a waste of time". We strongly disagree. Although these characters may be phylogenetically uninformative in the context of a specific study, they enhance the value of the matris for further use, when the addition of more taxa or changes in topology may render them informative. Whitlock is also correct that they are useful in diagnosing specimens.)
The present version of the manuscript has now also been submitted to PeerJ for formal peer review.