There are a number of possibilities. The printed index has two entries, as noted, for John and Henry, sharing a page reference which usually suggests twins.
Most likely is that there were twins and the entry for John has been badly transcribed or missed altogether when the GRO index was created and digitised.
Another option (less likely) is that it was one birth where the name was later changed from John to Henry.
Without getting certificates, you are always only working by assumption.