We have recently demonstrated, that DNA ejection from bacteriophage I can be partially or completely suppressed in vitro by external osmotic pressure. This suggests that DNA ejection from phage is driven by an internal mechanical force consisting of DNA bending and DNA-DNA electrostatic repulsion energies. In the present work we investigate the extent to which DNA ejection is incomplete at zero osmotic external pressure when phage is opened with its receptor in vitro. The DNA fragment remaining in the capsid and the tail that is no longer bent or compressed sand hence for which there is no internal driving force for ejections is shown not to be ejected. We also demonstrate that DNA can be "pulled" out from the capsid by DNase I acting as a DNA binding protein or spermine acting as a DNA condensing agent. In particular, cryo electron microscopy and gel electrophoresis experiments show the following: (i) DNA ejection from bacteriophage I incubated in vitro with its receptor is incomplete at zero external osmotic force, with several persistence lengths of DNA remaining inside the phage capsid, if no nuclease ( DNase I) or DNA condensing agent ( spermine) is present in the host solution; (ii) in the presence of both DNase I and spermine in the host solution, 60% (approximate to 29 kbp) of wild-type lambda DNA (48.5 kbp) remains unejected inside the phage capsid, in the form of an unconstrained toroidal condensate; (iii) with DNase I added, but no spermine, the ejection is complete; (iv) with spermine, but without DNase I added, all the DNA is again ejected, and organized as a toroidal condensate outside.