[R-SIG-Finance] subsetting based on date and time of an xts?

[Raghuraman Ramachandran]

Dear guRus

I have an xts variable res which is as below:

>str(res)
An ‘xts’ object on 2014-09-03 23:30:00/2014-10-01 21:45:00 containing:
  Data: num [1:1837, 1] 1991 1991 1991 1991 1991 ...
 - attr(*, "dimnames")=List of 2
  ..$ : NULL
  ..$ : chr "Data"
  Indexed by objects of class: [POSIXct,POSIXt] TZ:
  xts Attributes:
List of 4
 $ from   : chr "20140911  22:00:00"
 $ to     : chr "20141001  22:00:00"
 $ src    : chr "IB"
 $ updated: POSIXct[1:1], format: "2014-10-02 10:59:31.125498"

How do I build a new xts out of this which is a time series that
starts and ends at particular times everyday. As an example if I want
the data series to start at 2014-09-03 14:30:00 and end at 2014-09-03
21:15:00 and then continue with 2014-09-04 14:30:00 and end at
2014-09-04 21:15:00 and so on please?

Basically I am trying to rbind multilple subsets of the xts data based
on time everyday.

Many thanks for the kind assistance,
R