There is no doubt that the UNSM 93000 specimen of Nyctosaurus (Figs. 1, 2, Brown 1978, 1986) had three wing phalanges (not counting the hypothetical vestigial ungual perhaps still buried in the matrix, insufficiently excavated to reveal the possibility). All other pterosaurs have four wing phalanges, plus the ungual.

Figure 3. Cast of the UNSM93000 specimen of Nyctosaurus (hanging in my office and taken by cellphone). Wing phalanges marked. The identity of the wing phalanges appears to be m4.1, m4.2 and m4.4 (curved). There is no vestige or indication of m4.3. Arrows point to joints and the extensor tendon process.
Due to this specimen, traditional thinking holds that all Nyctosaurus specimens had but three wing phalanges, having lost the fourth phalanx. We’ll test that paradigm today.

Figure 2. The UNSM specimen of Nyctosaurus, the only one for which we are sure it had only three wing phalanges.
Contra traditional thinking
I found four wing phalanges on the more basal nyctosaurs like the Field Museum specimen FMNH 25026 (Fig. 3) and the Fort Hays specimen, FHSM VP21 (Fig. 4). The more derived crested specimen, KJ2 has a split and shattered wing phalanx 2/3, so I can’t determine its status from available data. Its reconstructed length and phylogenetic nesting strongly suggests that m4.3 is simply missing. It never developed.

Figure 3. The Field Museum Nyctosaurus, FMNHVP21, with color overlays identifying the wing phalanges of one wing. Four are present. Only the proximal knuckle and distal third of m4.3 is preserved.

Figure 4. The Fort Hays specimen of Nyctosaurus, FHSM VP2148 identifying all four wing phalanges and the extensor tendon process (ETP). The rostral crest is made of putty. Another unidentified curved m4.4 is beneath the distal end of the upper m4.1.
So what happened to the missing phalanx?
In the UNSM93000 specimen (Figs. 1, 2) m4.2 is roughly 85% of m4.1, which is similar in proportion to m4.2 in the other nyctosaurs. Manual 4.3 appears to be unreduced in the Field Museum specimen. Manual 4.3 is relatively shorter in the Fort Hays specimen, but still substantial. In all Nyctosaurus manual 4.4 is curved and shorter than the other phalanges, and that pattern in followed in UNSM 93000, Thus it does not appear that manual 4.4 withered and disappeared in UNSM 93000.
No, something else happened to m4.3.
Manual 4.3 does not appear to have fused to m4.2. That would have made an elongated wing phalanx and finger. There are no marks on the middle phalanx to that effect. Rather, all the other bones are much the same as in other nyctosaurs. It seems that m4.3 simply did not appear, did not develop. Very strange.
If the ungual of UNSM 93000 turns out to be missing (after further excavation) I would not be surprised. I’m sure it could have happened on the most derived Cretaceous pterosaurs. Especially one that was losing m4.3! A vestige can only last so long, but this specimen needs just a little bit more excavation to find out.
As always, I encourage readers to see specimens, make observations and come to your own conclusions. Test. Test. And test again.
Evidence and support in the form of nexus, pdf and jpeg files will be sent to all who request additional data.
References
Brown GW 1978. Preliminary report on an articulated specimen of Pteranodon Nyctosaurus) gracilis. Proceedings of the Nebraska Academy of Science 88: 39.
Brown GW 1986. Reassessment of Nyctosaurus: new wings for an old pterosaur. Proceedings of the Nebraska Academy of Science 96: 47.