I have to say as an editor and a collection manager, that I don't agree with the comment one person made that journals should 'require' (rather than 'recommend' like the ICZN Code does) that Types be deposited in an institution. This relies on an ideal-world
belief that 'all institutions' are better managed and more accessible than 'all private collections' - and this is simply (sadly) not true. Some institutes have a near-total embargo on loaning types and are totally 'inaccessible to all science' unless you
can afford to show up at the door - while some private collectors make their material very available. Think of Charlie and Lois O'Brien's private collection during their heyday, compared to certain major type-holding museums- and which you would rather consult
a specimen from...?
I am editor in a journal that has a written policy against type deposition in private collections, and we lost a very good paper recently because a German expert did not want to deposit a unique holotype specimen away from his reference collection, because
he was continuing to work on the group and did not want his own access to the taxon he described to be curtailed in future- his collection will pass to a major museum on his death, like Charlie's did, and he attends many international meetings, so the accessibility
of the type (which was of course published in a different journal) is as good as most museums and better than some- so the only consequence of the 'rule' was that one journal shot itself in the foot and lost a good paper and some goodwill!
The life-span of a private collector is of course a tiny proportion of the life-span of a specimen, and most sensible collectors make arrangements for their collection to pass to what they judge to be a good institution, so in the end it is all the same. Of
course many private individuals prefer to deposit their holotypes right away, and that is good (we gratefully received several at NHM this year, and they are well cared for and accessible) but it shouldn't be because a journal demands it.